Touch and Petting: How Dogs Respond to Being Handled

1. Why Touch Deserves a Closer Look

1.1 Touch Is Not One Thing

"Petting" sounds like a single behavior, but the contact people have with dogs covers very different situations. A dog that pushes its head under its owner's hand on the sofa, a dog that is stroked by a stranger in the street, a dog whose ears are cleaned and a dog that is held still for a blood sample are all being touched. From the dog's point of view, these situations differ in who starts the contact, whether the dog can end it, how predictable it is and what it usually leads to.

For this article it is useful to separate three broad kinds of touch. Affiliative touch is social contact that has no purpose beyond the interaction itself: stroking, scratching, leaning together. Handling is touch with a practical goal, such as checking paws, brushing, drying, examining or medicating. Restraint is handling in which movement is deliberately limited, from a hand on the chest to a full-body hold. The categories overlap, and the same hand on the same shoulder can be affiliative in one moment and restraint in the next if the dog tries to leave and is prevented from doing so.

The research reflects these differences. Studies of affiliative touch tend to measure preference and hormonal changes; studies of handling and restraint tend to measure fear, escape attempts and heart rate.

1.2 What the Research Can and Cannot Answer

Three kinds of questions run through the studies discussed here. The first is whether dogs value petting: will they work for it, choose it over alternatives, or spend time near a person who provides it? The second is what happens in the body during and after contact: do hormones or heart rate change, and in which direction? The third is how dogs behave while they are being touched: do they lean in, stay neutral, freeze, or show signs that are usually read as conflict or appeasement?

None of these questions measures the dog's experience directly. A preference shows relative value under particular conditions, a hormone level shows a physiological change that can accompany different emotional states, and a behavior shows what the dog does, not what it feels (what stress measures in dogs can and cannot show). The most useful conclusions come where several kinds of evidence point in the same direction.

2. Petting as a Reinforcer

2.1 Food Versus Brief Social Interaction

The most systematic tests of petting as a consequence come from a series of experiments by Feuerbacher and Wynne. In the first, an arbitrary response (touching a target with the nose) was reinforced either with a small piece of food or with brief human social interaction. Shelter dogs, owned dogs and hand-reared wolves were tested. Across all three populations, brief social interaction was a relatively ineffective reinforcer compared with food for most animals, with lower response rates and longer latencies (Feuerbacher & Wynne, 2012).

The qualifiers matter. The interaction lasted only seconds, and the response was an arbitrary one chosen for experimental control. The study does not show that dogs do not value people. It shows that a few seconds of social interaction maintained less responding than a small food item in most individuals under these conditions.

2.2 Food or Petting: A Choice That Shifts

A second study offered owned dogs a direct choice between food and petting (Feuerbacher & Wynne, 2014). Dogs completed five-minute sessions in which petting was continuously available from one person while food was available at another location, first continuously and then on progressively thinner schedules, then not at all, and finally again as at the start. Dogs were tested either in a familiar daycare setting or in an unfamiliar laboratory room, and the petting was delivered either by the owner or by a stranger.

When food was readily available, the dogs preferred food to petting, as the title of the paper states. As food became scarcer, all groups spent less time at the food location. The time dogs spent with the petting person varied with the setting, with how familiar the person was, and with how deprived the dogs had been of social interaction. The value of petting, in other words, was not a fixed quantity. It shifted with what else was available, with the place and with the person (how the value of food, play and social rewards is measured).

2.3 Petting Versus Praise

If food and petting differ, different forms of social interaction differ as well. In a further set of experiments, dogs preferred petting to vocal praise, both when the two were available at the same time from different people and when only one type of interaction was available in a session (Feuerbacher & Wynne, 2015). The study measured choice, not training performance, so it does not show how much petting improves learning. But it does indicate that, for the dogs tested, physical contact had more value than words alone.

2.4 Who Pets Matters

Another study kept the type of petting constant and varied the person. Two people provided identical petting, and dogs could choose where to spend their time (Feuerbacher & Wynne, 2017). Owned dogs preferred their owners in an unfamiliar environment but spent more time with the stranger in a familiar one. Shelter dogs, and owned dogs tested with two strangers, quickly formed strong preferences for one stranger over the other, with preference magnitudes comparable to the owned dogs' preference for their owner in the unfamiliar setting.

This result corrects two common assumptions: that dogs always want contact with their owner more than with anyone else, and that a stranger's touch is always less welcome. Both depended on context. One possible interpretation, not tested directly, is that the owner provides security in an unfamiliar place.

2.5 What This Means for Training

Taken together, these studies show that petting can function as a reinforcer, that it is often weaker than food when food is easily available, that it is preferred to praise in choice tests, and that its value depends on the setting and the person. For training, the implication is not that petting is useless, but that it cannot be assumed to be rewarding at any moment. A dog that is excited, hungry, focused on a task or in an unfamiliar place may not value a pat on the head as much as the person giving it believes.

3. What Happens in the Body

3.1 Early Neurochemical Data

Much of the popular belief that petting is good for dogs rests on hormonal studies. An early study measured several neurochemicals in people and dogs before and after a period of positive interaction (Odendaal & Meintjes, 2003). Concentrations of beta-endorphin, oxytocin, prolactin, beta-phenylethylamine and dopamine increased in both species after the interaction, while cortisol decreased in the humans only.

The cortisol decrease, often quoted as evidence that petting relaxes dogs, was reported for the people, not for the dogs. The study was exploratory and is best read as an early indication that affiliative interaction is associated with changes in several systems in both species.

3.2 Oxytocin and Cortisol During Stroking

A later exploratory study used intravenous catheters in dog owners and their dogs, which allowed repeated blood sampling before, during and after a short interaction (Handlin et al., 2011). The participants were ten female owners and their male Labrador Retrievers, and the interaction took place during the first three minutes of a one-hour sampling period. Oxytocin rose in the dogs within the first few minutes of gentle stroking.

Cortisol did not behave as many readers would expect. The owners' cortisol decreased, but the dogs' cortisol increased over the session (Petersson et al., 2017). In the dogs, the negative correlation between oxytocin and cortisol was r = −0.49 with p = 0.057, a trend rather than a statistically significant result (Handlin et al., 2011). The study therefore supports the claim that stroking by a familiar person is associated with rising oxytocin in dogs, but it does not show that stroking lowered the dogs' stress hormone levels in this setting.

A follow-up analysis of the video recordings from the same experiment examined how the owners actually touched their dogs (Petersson et al., 2017). Touch was coded as stroking, scratching or patting, and scratching and patting together were grouped as "activating touch". The more often owners used activating touch, the higher the dogs' cortisol was 15 minutes later (Rs = 0.661, p = 0.038). The lower the dogs' oxytocin during the interaction, the more stroking they received (Rs = −0.717, p = 0.041), and owners with low oxytocin levels touched their dogs more often.

These are correlations in ten pairs, computed across many variables, and the authors describe the study as exploratory. They do not show that patting causes a stress response. They do suggest that "petting" is too broad a category, and that the type and rhythm of touch may matter as much as whether a dog is touched at all. The wider oxytocin literature is reviewed separately (what oxytocin research in dogs does and does not show).

3.3 Physical Contact at Reunion

One study isolated the effect of physical contact more clearly. Twelve female Beagles were each left alone in a familiar test arena and then reunited with a familiar person under three conditions, presented in a balanced order: the person initiated calm physical and verbal contact, the person talked to the dog without touching it, or the person ignored the dog (Rehn et al., 2014). Each interaction lasted four minutes, and blood was sampled for oxytocin and cortisol.

Oxytocin rose on the person's return in all conditions, which the authors attributed to the dogs seeing the person approach. Only when physical contact was included did oxytocin stay elevated after the interaction had ended. Cortisol declined across the test in all conditions, but most markedly with physical contact. Dogs in the physical-contact condition also initiated more physical contact with the person and showed more lip licking at reunion. When only verbal contact was given, the first responses were tail wagging and vocalizing, and when the dogs were ignored, they may have redirected their approach toward an assistant who was always present in the room.

This is one of the clearer experimental findings in the area, but its population is narrow: twelve laboratory Beagles, all female, with a familiar person who was not their owner. It shows that in this setting physical contact added something that the mere return of the person did not. It does not show that every dog benefits from being touched at every reunion, and the increase in lip licking is a reminder that a dog's behavior during welcome contact can include signals that are often read as stress or appeasement.

3.4 How Firm Are These Findings?

A review of oxytocin and human–dog interaction concluded that the field is characterized by small samples, heterogeneous methods and differences between assays (Powell et al., 2019). The studies above fit that description: the hormonal experiments worked with ten or twelve dogs, the populations are specific, and the hormone measures depend on sampling method and analysis.

The defensible summary is moderate. Calm physical contact with a familiar person has been associated with rising oxytocin in dogs, and in one experiment physical contact at reunion was followed by a more pronounced cortisol decline. In another, the dogs' cortisol rose during owner interaction. The physiology of petting is real, but neither uniform nor as well established as popular summaries suggest.

3.5 After the Interaction

Social contact may also affect what happens after it ends. In a within-subject study, pet dogs either experienced a positive social interaction (petting and ball play) or a negative one (separation, a threatening approach and a still-face test) before a monitored nap (Kis et al., 2017). After the negative experience, dogs fell asleep faster, spent less time drowsy and more time in REM sleep than after the positive one. The positive condition combined petting with play, so the study does not isolate touch, but it shows that the emotional quality of a social interaction can be detected in subsequent sleep (how experiences before sleep change sleep in dogs).

4. Contact in Stressful Settings

4.1 Shelter Studies

Some of the strongest evidence that human contact can reduce physiological stress markers comes from shelters, where dogs are separated from familiar people and placed in a novel, noisy environment. In an early study, shelter dogs underwent a blood sampling procedure and were then either petted or left unhandled. Dogs that were not petted showed elevated cortisol, whereas dogs petted by either a male or a female petter did not; petted dogs also showed relaxed postures and fewer escape attempts (Hennessy et al., 1998).

In a later study, stray dogs on their second day in a shelter either received a 45-minute human contact session, which combined play, grooming, petting and a review of basic obedience, or no session. Dogs that had the contact session had lower cortisol levels on day three than dogs that did not; breed type, sex and age had no effect on cortisol on any day measured (Coppola et al., 2006). The effect was reported for the day after the session, not for the whole observation period.

A third study compared different kinds of human contact in newly admitted shelter dogs. Thirty-minute sessions with a person who was passively present, petted the dog or played with it were associated with lower plasma cortisol than control conditions, along with less vocalizing and panting, and passive presence produced effects comparable to some of the active conditions (Shiverdecker et al., 2013).

The last finding is easy to overlook. In this study, a person sitting quietly with the dog did as much for cortisol as some forms of active contact. Petting was one way to provide human company, but not the only one. The broader research on social buffering is covered separately (how the presence of people buffers stress in dogs).

4.2 Owner Contact During a Veterinary Examination

A veterinary examination is another situation in which dogs are under stress and are touched by someone other than their owner. In a crossover study, owners either petted and talked to their dog during a standardized examination or were present without interacting (Csoltova et al., 2017). The examination produced acute stress responses in both conditions. When owners interacted with their dogs, the dogs made fewer attempts to jump off the table, and their heart rate and maximal ocular surface temperature were attenuated.

Owner contact reduced some indicators of stress but did not remove them, and it combined petting with talking. It suggests that comforting contact from a familiar person can be a useful part of veterinary handling, not that it makes an examination pleasant.

4.3 A Familiar Person as a Safe Haven

A study of thirty dogs examined the effect of the owner's presence on the response to a threatening stranger (Gácsi et al., 2013). The owner's presence dampened the dogs' heart-rate response to the threat, and dogs that first met the threatening stranger with the owner present showed a smaller heart-rate increase when they later met the same stranger alone. The study measured presence, not touch, but it is relevant because it shows that much of the comfort a familiar person provides may come from being there, not necessarily from physical contact.

4.4 The Transfer Problem

Shelter dogs, laboratory Beagles and dogs on an examination table are study populations with high baseline stress and unfamiliar surroundings, in which almost any friendly human contact has room to make a difference. A family dog lying relaxed on the sofa starts from a very different position, and the same stroking may add little or even be an interruption. The shelter results show that human contact can reduce physiological stress markers in stressed dogs, not that more petting is always better for dogs that are not stressed.

5. Body Region and Type of Touch

5.1 Chest and Head

The question of where dogs like to be touched has been examined in one group of studies with heart rate measurements. Twenty-eight privately owned dogs experienced nine different physical interactions while their behavior and cardiovascular parameters were recorded (Kuhne et al., 2014a). Petting and holding the dog around the head was associated with increased SDNN, a measure of overall heart rate variability, while petting at the chest was associated with increased vagal tone. Displacement activities correlated negatively with all cardiovascular parameters, and appeasement gestures correlated positively with heart rate.

The authors concluded that some common forms of physical human–dog contact were perceived as unpleasant by the dogs and called for guidelines on pleasant and safe contact. Heart rate variability is not a simple measure of comfort, and its interpretation depends on the context and on which component is measured. The study does not provide a ranking of body regions. What it does show is that the region and type of contact were associated with different physiological and behavioral responses, and that the head, the place where many people automatically reach, was not the region associated with the clearest vagal response.

5.2 Activating and Calm Touch

The video analysis from the oxytocin experiment points in a similar direction from a different angle: more frequent scratching and patting by owners went with higher cortisol in their dogs 15 minutes later (Petersson et al., 2017). This is a correlation in a very small sample, and it cannot separate the touch from the owner's general style or from the dog's behavior that might have prompted it. It fits, however, with the idea that the quality of touch matters, not just its presence.

5.3 What Is Not Known

Popular advice often names the chest, shoulders and base of the ears as preferred areas and the top of the head, paws and tail as areas to avoid. The data reviewed here are compatible with that advice for the chest and the head, but no study reviewed here has systematically compared all body regions in a large sample of companion dogs. Individual preferences, breed differences, age and previous experiences, including painful ones, have not been mapped. A region that most dogs tolerate may be the region that one particular dog avoids.

In practice: A dog that suddenly starts to avoid touch in one area, or reacts to contact where it used to accept it, is a reason for a veterinary check before it is treated as a training issue. This comes from practice, not from the studies cited here; the relationship between pain and changes in behavior is covered separately (why dogs often hide pain).

6. Familiarity

6.1 Familiar and Unfamiliar Handlers

The same research group also asked whether it matters who does the touching. Twenty-four privately owned dogs were exposed to nine different tactile interactions of 30 seconds each, among them holding a paw and stroking the head, performed by either a familiar or an unfamiliar person (Kuhne et al., 2012). Familiarity significantly influenced how often the dogs initiated redirected behaviors (F1,184 = 4.94, p = 0.027) and the frequency (F1,193 = 10.67, p = 0.001) and duration of appeasement gestures. The authors suggested that these behaviors may indicate the quality of the relationship between dog and person.

6.2 Familiarity and Heart Rate

A further study with 24 dogs combined the behavioral coding with heart rate and heart rate variability (Kuhne et al., 2014b). The nine kinds of contact included petting and holding the lying dog on the ground and covering the dog's muzzle with one hand. Physiological responses differed between the familiar and the unfamiliar person conditions, and appeasement gestures and redirected behaviors occurred together with physiological signs of stress. The authors argued that combining behavioral and physiological measures can show whether a form of contact creates emotional conflict for the dog, which would undermine the benefits that human–dog contact is supposed to bring.

Some of the interactions in these studies, such as covering the muzzle or holding a lying dog on the ground, resemble handling or restraint more than petting. Their inclusion shows a continuum rather than two separate categories.

6.3 Fitting the Findings Together

At first sight, the familiarity findings seem to conflict with the choice studies, in which dogs sometimes preferred a stranger's petting to their owner's (Feuerbacher & Wynne, 2017). The studies measured different things. In the choice studies, the dog decided whether to approach and stay; nothing was imposed. In the Kuhne studies, the interactions were applied for a fixed time regardless of the dog's response. The combined picture is that a familiar person is not automatically preferred, and that the same imposed contact is not experienced in the same way regardless of who provides it. Neither line of research has tested both questions in the same dogs.

7. Stillness Is Not Consent

7.1 Behavioral Suppression During Grooming

A dog that stands still while it is touched is often described as "fine with it". One observational study challenges this reading directly. Across the phases of a routine grooming appointment, heart rate variability declined from arrival through restraint, nail trimming, bathing and drying, with the lowest values during drying, while behavioral suppression increased (Hart & Hart, 2026). Suppression was inversely related to heart rate variability. The authors' interpretation is that stillness during grooming may reflect coping rather than comfort. The study was published online only in 2026 and should be read as a recent single finding, not as an established, widely replicated result.

The study is observational, so it cannot separate the effect of each grooming step from the passage of time, and the grooming setting combines handling with noise, water and an unfamiliar environment. But the pattern is important beyond grooming: a dog that does less while being handled is not necessarily a dog that feels better. The concept of learned passivity in aversive situations is covered separately (what learned helplessness means in dogs).

7.2 Subtle Signals That People Miss

If stillness is unreliable, attention shifts to smaller behaviors. In a questionnaire study of 1,190 owners, the behaviors most often recognized as signs of stress were obvious ones such as trembling and whining. Subtler behaviors such as looking away, turning the head, yawning and nose licking were named much more rarely (Mariti et al., 2012). These are exactly the behaviors that a dog may show while it is being petted.

Caution is needed in the other direction as well. When dogs were approached by an unfamiliar person in a threatening and a neutral manner, most displacement behaviors were not performed more often in the threatening condition. Blinking, nose licking and lip wiping went with a non-reactive attitude in both conditions, and only paw lifting lasted longer under the threatening approach (Pedretti et al., 2023). A single lip lick during petting is therefore not proof of distress. In the reunion study, dogs that received physical contact showed more lip licking while also seeking more contact (Rehn et al., 2014). Signals have to be read in clusters and in context (how to read canine body language in context).

7.3 Conflict During Contact

The studies on body region and familiarity found appeasement gestures and redirected behaviors more often in some contact conditions than in others, and these behaviors were associated with physiological signs of stress (Kuhne et al., 2012, 2014a, 2014b). A dog can want to be near a person and at the same time find a particular form of contact uncomfortable. This mixture of approach and avoidance is not unusual, and it explains why dogs sometimes solicit contact and then turn away from it (what approach-avoidance conflict looks like in dogs).

7.4 Choice as the Better Measure

The choice studies offer a way out of interpreting single signals. When the dog can move away and comes back, its behavior says something about the value of the contact (Feuerbacher & Wynne, 2014, 2017). When it cannot, the grooming data show how misleading the absence of movement can be (Hart & Hart, 2026) (why control and predictability matter for stress).

8. Handling and Restraint

8.1 Fear at the Veterinary Practice

Handling is the form of touch that dogs most often cannot avoid. In a study of 135 dogs at a veterinary practice, up to 78.5 percent were classified as fearful on the examination table, fewer than half entered the clinic calmly, and 13.3 percent had to be dragged or carried in (Döring et al., 2009). For many dogs, being handled by unfamiliar people in an unfamiliar place is not a neutral event.

8.2 Restraint Technique

Whether the way a dog is held makes a difference has been tested directly. In a randomized block design with 97 shelter dogs, five conditions were compared during a two-minute examination: passive restraint, muzzle-hold restraint, a basket muzzle, a soft muzzle and full-body restraint (Cisneros et al., 2025). Fear scores were lower under passive restraint than under each of the other four conditions, and dogs in the muzzle-hold condition made more escape attempts. Physiological measures did not differ between conditions.

The study used shelter dogs in a short examination, so it does not tell us how family dogs respond to a full veterinary visit. It does show that the degree of restriction changes behavioral responses even when the procedure is the same. The muzzle hold resembles one of the interactions in the familiarity studies, covering the dog's muzzle with one hand (Kuhne et al., 2014b), a gesture people use casually.

8.3 Handling Is a Skill the Dog Can Learn

The difference between affiliative touch and handling is not only in the hand but in the dog's expectations. Handling that has repeatedly led to discomfort can make a dog wary of all touch in the same region, and touch in that region may then trigger anticipation of the procedure. Conversely, handling that is predictable, brief and paired with something the dog values can become more acceptable. The research on structured handling training, often called cooperative care, is covered separately (what cooperative care can and cannot achieve), as is the learning method behind it (how desensitization and counterconditioning work).

In practice: We keep affiliative touch and handling training separate. Stroking stays something the dog can ask for and end, while handling is practiced in short, structured sessions with a clear start signal, a food reward and the option for the dog to step away. This separation comes from practice; it has not been tested as such in a controlled study.

9. Strangers and Greetings

9.1 Friendly and Threatening Approaches

Most encounters between dogs and strangers begin with an approach, and how the approach is made affects how the dog responds. In one study, adult pet dogs faced an unfamiliar woman who approached them in a friendly manner (walking normally, talking and petting the dog) and in a threatening manner (approaching slowly with a steady gaze and without speaking) (Vas et al., 2005). In the first experiment, 30 dogs of 19 breeds responded appropriately to both kinds of approach, and the order of the two episodes had no significant effect. In the second, 60 dogs from three breed groups experienced a friendly greeting followed by a threatening approach. There were significant breed-group differences in how dogs responded when the person's behavior switched, with Belgian Shepherds more often responding with avoidance or aggression than the other groups.

Dogs read the body language of an approaching stranger and respond to it, and they differ in how sensitive they are to such cues. The threatening approach was a deliberately exaggerated combination of slowness, staring and silence; everyday greetings are rarely so clear.

9.2 What Has Not Been Measured

Common advice says that strangers should not reach over a dog's head, should not loom over a dog, and should let the dog approach first. These recommendations are plausible in the light of the studies above: the head was not the region associated with the clearest vagal response in one study (Kuhne et al., 2014a), the familiarity of the person touching the dog influenced appeasement and redirected behavior (Kuhne et al., 2012), and dogs responded to the manner of a stranger's approach (Vas et al., 2005). No study reviewed here, however, has directly compared reaching over the head with other forms of greeting. The advice is reasonable, but it is an inference, not a measured result.

9.3 Greeting Guidance

In practice: Our standard advice for greeting an unfamiliar dog is to ask the owner first, stand or crouch slightly sideways rather than bending over the dog, let the dog come to you, and if it does, touch it briefly on the side of the chest or shoulder rather than on top of the head. After a few seconds, stop and see whether the dog stays or comes back for more. This is guidance from practice, built on the studies above but not tested as a complete procedure.

10. Children and Touch

10.1 What Children Actually Do

Children touch dogs differently from adults: more often, more impulsively and with more whole-body contact. In a survey of 402 caregivers of children up to six years old, hugging the dog was frequently observed by 46 percent of caregivers and kissing it by 27 percent (Arhant et al., 2017). These are interactions that took place in the presence of an adult, who did not consider them a problem.

No peer-reviewed study reviewed here measured how dogs respond physiologically to being hugged. What can be said is that a hug combines close contact around the head and neck with restriction of movement. In the studies on contact, head contact was associated with a different physiological response from chest contact (Kuhne et al., 2014a), and more restrictive handling produced higher fear scores than passive restraint (Cisneros et al., 2025).

10.2 Touch Before Bites

An analysis of publicly posted videos of dog bites examined what happened in the seconds before the bite. Of 143 videos, 56 were coded second by second. Non-neutral body postures and some displacement and appeasement behaviors increased from about 20 seconds before the bite, and human tactile contact with the dog increased from about 21 seconds before (Owczarczak-Garstecka et al., 2018). The study reports no universal sequence, and not every behavior described as an appeasement signal showed a clear time pattern. But touch was part of the run-up to many of the recorded bites. The broader evidence on bites involving children is reviewed separately (what the research shows about dog bites and children).

10.3 Teaching Children to Read Dogs

Children often misread dogs. Young children frequently interpreted aggressive dog faces as happy, which was the most common error in one study (Meints & Racca, 2026). Training can help: in a study of 124 children aged three to five and 40 parents, participants were taught to identify dog signals from video, and 88 children completed the full longitudinal protocol (Meints et al., 2018). Teaching children what a dog looks like when it wants to be left alone is one of the few touch-related interventions that has been studied systematically.

In practice: We teach children a simple rule: they may stroke a dog only when an adult says yes and the dog comes to them, with one hand on the side of the body, for as long as it takes to count to three, and then they stop and see whether the dog stays. Hugs and kisses are not part of the rule. This comes from practice and has not been evaluated in a study.

11. What Follows for Practice

11.1 The Consent Test

The most useful practical consequence of the research is a change of question. Instead of asking whether a dog likes being petted, it is more productive to ask whether this dog wants this kind of contact from this person right now. The choice studies suggest how to find out: give the dog the opportunity to choose (Feuerbacher & Wynne, 2014, 2017).

In practice: A simple consent test works like this. Stroke the dog for a few seconds, then take your hand away and wait. If the dog moves closer, nudges your hand or leans in, it is asking for more. If it stays where it is without seeking contact, shakes off, turns away or moves away, the contact was not wanted, or not wanted now. The test is a tool from practice, not a validated instrument, but it follows the logic of the preference studies: the dog's own behavior when it is free to choose is more informative than its behavior while it is being touched.

11.2 Reading the Individual Dog

The studies describe groups. Every one of them reports individual differences, and none can tell a reader how a particular dog responds. Age, health, breed, previous experience, the current level of arousal and the accumulated stress of the day all affect how a touch is received (how stress accumulates across the day). The consent test, the separation of affiliative touch and handling, and attention to the dog's behavior when it is free to leave are ways of applying the group findings to the individual.

12. Summary at a Glance

• Petting is a reinforcer with variable value: brief social interaction was weaker than food for most dogs, food was preferred to petting when it was readily available, and petting was preferred to praise (Feuerbacher & Wynne, 2012, 2014, 2015).

• The person and the place matter: owned dogs preferred their owner's petting in an unfamiliar place but not in a familiar one, and some dogs quickly preferred one stranger over another (Feuerbacher & Wynne, 2017).

• Hormonal data are suggestive: oxytocin rose during stroking, but dogs' cortisol rose in one study and fell most with physical contact in another (Handlin et al., 2011; Petersson et al., 2017; Rehn et al., 2014).

• Shelter dogs benefited from human contact: petting, play and passive presence were associated with lower cortisol (Hennessy et al., 1998; Coppola et al., 2006; Shiverdecker et al., 2013).

• Body region and familiarity changed responses: chest and head contact were associated with different heart rate measures, and familiarity influenced appeasement and redirected behavior (Kuhne et al., 2012, 2014a, 2014b).

• Stillness can mean coping: behavioral suppression during grooming was associated with lower heart rate variability (Hart & Hart, 2026).

• Restraint matters: passive restraint produced lower fear scores than more restrictive methods (Cisneros et al., 2025).

• Children and touch: hugging is common in families, and human tactile contact increased in the seconds before many recorded bites (Arhant et al., 2017; Owczarczak-Garstecka et al., 2018).

13. Research Gaps and Critical Appraisal

13.1 Small Samples

Most of the studies on touch are small. The hormonal experiments worked with ten owner–dog pairs (Handlin et al., 2011; Petersson et al., 2017) and twelve Beagles (Rehn et al., 2014). The body-region and familiarity studies included 24 or 28 dogs (Kuhne et al., 2012, 2014a, 2014b). The hormonal, body-region and familiarity findings therefore each rest on a few dozen dogs at most, spread across very different designs. Effects found in such samples can be unstable, and the absence of an effect may simply reflect too little statistical power.

13.2 Specific Populations

The populations are not representative of family dogs. The reunion study used laboratory Beagles, all female (Rehn et al., 2014). The oxytocin study used male Labradors with female owners (Handlin et al., 2011). The cortisol studies with clear reductions came from shelters, where dogs were isolated and stressed (Hennessy et al., 1998; Coppola et al., 2006; Shiverdecker et al., 2013), and the restraint study also used shelter dogs (Cisneros et al., 2025). How far these results apply to relaxed family dogs at home has not been tested directly.

13.3 Ambiguous Measures

Oxytocin, cortisol and heart rate variability can all change with arousal of either valence. A rise in oxytocin does not prove pleasure, and a rise in cortisol does not prove distress. The reviews of the oxytocin literature point to differences in assays and sampling methods (Powell et al., 2019). Behavioral measures have their own problems: displacement behaviors such as lip licking were not specific to threatening situations in one study (Pedretti et al., 2023) and appeared alongside contact seeking in another (Rehn et al., 2014).

13.4 Imposed Versus Chosen Contact

Two traditions rarely meet. The choice studies let the dog decide but measure only time allocation and responding (Feuerbacher & Wynne, 2014, 2017). The physiological studies impose standardized contact and measure the body's response (Kuhne et al., 2014a; Rehn et al., 2014). No study reviewed here has combined free choice with physiological measurement in the same dogs, which is exactly what would be needed to tell whether the contact dogs choose is also the contact that relaxes them.

13.5 Unanswered Practical Questions

Many everyday questions remain unstudied: how dogs respond to hugs, whether reaching over the head is worse than approaching from the side, how long petting should last, how age and pain change the response, and whether consent tests predict welfare outcomes. Practical advice in these areas, including the advice in this article, rests on inference and experience.

13.6 Observational Designs

Some of the most practically relevant findings come from observational designs. The grooming study could not assign dogs to different handling methods (Hart & Hart, 2026), the bite video study depended on what people chose to film and post (Owczarczak-Garstecka et al., 2018), and the survey of caregivers relied on reports (Arhant et al., 2017). These studies identify patterns worth testing, not causal effects.

14. Conclusion

The research does not support the simple belief that dogs love being petted, nor the opposite claim that dogs merely tolerate it. Petting can be a reinforcer and can be preferred to praise. Calm contact with a familiar person has been associated with hormonal changes compatible with relaxation, and human contact was associated with lower cortisol in stressed shelter dogs. At the same time, the value of petting depended on alternatives, place and person, the region and type of touch changed the dog's response, restrictive handling increased fear, and stillness during handling accompanied physiological signs of stress.

The practical conclusion is that touch is something to offer, not something to apply. A dog that can approach, stay or leave shows by its own behavior whether a particular contact is welcome. Asking that question repeatedly, and separating affiliative touch from handling that the dog has to learn to accept, is a better use of the evidence than any rule about where and how long to pet.

Key Insights (Takeaways)

• Petting can function as a reinforcer, but most dogs preferred food when it was readily available, and the value of petting shifted with place and person (Feuerbacher & Wynne, 2014, 2017).

• Calm physical contact with a familiar person was associated with rising oxytocin, but in one study the dogs' cortisol rose during owner interaction (Handlin et al., 2011; Petersson et al., 2017).

• Physical contact at reunion kept oxytocin elevated and was followed by the most marked cortisol decline in twelve laboratory Beagles (Rehn et al., 2014).

• Human contact, including passive presence, was associated with lower cortisol in stressed shelter dogs; this does not show that more petting is always better for relaxed family dogs (Shiverdecker et al., 2013).

• Contact at the chest and around the head was associated with different heart rate responses, and familiarity affected appeasement and redirected behavior (Kuhne et al., 2012, 2014a).

• A dog that stands still is not necessarily comfortable: behavioral suppression during grooming went with lower heart rate variability (Hart & Hart, 2026).

• Restrictive restraint produced higher fear scores than passive restraint (Cisneros et al., 2025).

• The most informative test is the dog's choice: pet briefly, pause and see whether the dog asks for more.

References

Arhant, C., Beetz, A. M., & Troxler, J. (2017). Caregiver reports of interactions between children up to 6 years and their family dog – implications for dog bite prevention. Frontiers in Veterinary Science, 4, 130. https://doi.org/10.3389/fvets.2017.00130

Cisneros, A., Carroll, A. D., Moody, C. M., & Stellato, A. C. (2025). Handle with care: Dogs show negative responses to restrictive handling restraints and tools during routine examinations. Applied Animal Behaviour Science, 286, 106601. https://doi.org/10.1016/j.applanim.2025.106601

Coppola, C. L., Grandin, T., & Enns, R. M. (2006). Human interaction and cortisol: Can human contact reduce stress for shelter dogs? Physiology & Behavior, 87(3), 537–541. https://doi.org/10.1016/j.physbeh.2005.12.001

Csoltova, E., Martineau, M., Boissy, A., & Gilbert, C. (2017). Behavioral and physiological reactions in dogs to a veterinary examination: Owner-dog interactions improve canine well-being. Physiology & Behavior, 177, 270–281. https://doi.org/10.1016/j.physbeh.2017.05.013

Döring, D., Roscher, A., Scheipl, F., Küchenhoff, H., & Erhard, M. H. (2009). Fear-related behaviour of dogs in veterinary practice. The Veterinary Journal, 182(1), 38–43. https://doi.org/10.1016/j.tvjl.2008.05.006

Feuerbacher, E. N., & Wynne, C. D. L. (2012). Relative efficacy of human social interaction and food as reinforcers for domestic dogs and hand-reared wolves. Journal of the Experimental Analysis of Behavior, 98(1), 105–129. https://doi.org/10.1901/jeab.2012.98-105

Feuerbacher, E. N., & Wynne, C. D. L. (2014). Most domestic dogs (Canis lupus familiaris) prefer food to petting: Population, context, and schedule effects in concurrent choice. Journal of the Experimental Analysis of Behavior, 101(3), 385–405. https://doi.org/10.1002/jeab.81

Feuerbacher, E. N., & Wynne, C. D. L. (2015). Shut up and pet me! Domestic dogs (Canis lupus familiaris) prefer petting to vocal praise in concurrent and single-alternative choice procedures. Behavioural Processes, 110, 47–59. https://doi.org/10.1016/j.beproc.2014.08.019

Feuerbacher, E. N., & Wynne, C. D. L. (2017). Dogs don't always prefer their owners and can quickly form strong preferences for certain strangers over others. Journal of the Experimental Analysis of Behavior, 108(3), 305–317. https://doi.org/10.1002/jeab.280

Gácsi, M., Maros, K., Sernkvist, S., Faragó, T., & Miklósi, Á. (2013). Human analogue safe haven effect of the owner: Behavioural and heart rate response to stressful social stimuli in dogs. PLOS ONE, 8(3), e58475. https://doi.org/10.1371/journal.pone.0058475

Handlin, L., Hydbring-Sandberg, E., Nilsson, A., Ejdebäck, M., Jansson, A., & Uvnäs-Moberg, K. (2011). Short-term interaction between dogs and their owners: Effects on oxytocin, cortisol, insulin and heart rate – An exploratory study. Anthrozoös, 24(3), 301–315. https://doi.org/10.2752/175303711X13045914865385

Hart, D. S., & Hart, S. J. (2026). Behavioural suppression and autonomic stress responses in dogs during routine grooming procedures. Applied Animal Behaviour Science, advance online publication, Article 107146. https://doi.org/10.1016/j.applanim.2026.107146

Hennessy, M. B., Williams, M. T., Miller, D. D., Douglas, C. W., & Voith, V. L. (1998). Influence of male and female petters on plasma cortisol and behaviour: Can human interaction reduce the stress of dogs in a public animal shelter? Applied Animal Behaviour Science, 61(1), 63–77. https://doi.org/10.1016/S0168-1591(98)00179-8

Kis, A., Gergely, A., Galambos, Á., Abdai, J., Gombos, F., Bódizs, R., & Topál, J. (2017). Sleep macrostructure is modulated by positive and negative social experience in adult pet dogs. Proceedings of the Royal Society B: Biological Sciences, 284(1865), 20171883. https://doi.org/10.1098/rspb.2017.1883

Kuhne, F., Hößler, J. C., & Struwe, R. (2012). Effects of human–dog familiarity on dogs' behavioural responses to petting. Applied Animal Behaviour Science, 142(3–4), 176–181. https://doi.org/10.1016/j.applanim.2012.10.003

Kuhne, F., Hößler, J. C., & Struwe, R. (2014a). Behavioral and cardiac responses by dogs to physical human–dog contact. Journal of Veterinary Behavior, 9(3), 93–97. https://doi.org/10.1016/j.jveb.2014.02.006

Kuhne, F., Hößler, J. C., & Struwe, R. (2014b). Emotions in dogs being petted by a familiar or unfamiliar person: Validating behavioural indicators of emotional states using heart rate variability. Applied Animal Behaviour Science, 161, 113–120. https://doi.org/10.1016/j.applanim.2014.09.020

Mariti, C., Gazzano, A., Moore, J. L., Baragli, P., Chelli, L., & Sighieri, C. (2012). Perception of dogs' stress by their owners. Journal of Veterinary Behavior, 7(4), 213–219. https://doi.org/10.1016/j.jveb.2011.09.004

Meints, K., Brelsford, V., & De Keuster, T. (2018). Teaching children and parents to understand dog signaling. Frontiers in Veterinary Science, 5, 257. https://doi.org/10.3389/fvets.2018.00257

Meints, K., & Racca, A. (2026). Look, he's smiling! Children's misinterpretation of dogs' facial expressions. Anthrozoös, 39(2), 315–342. https://doi.org/10.1080/08927936.2025.2551434

Odendaal, J. S. J., & Meintjes, R. A. (2003). Neurophysiological correlates of affiliative behaviour between humans and dogs. The Veterinary Journal, 165(3), 296–301. https://doi.org/10.1016/S1090-0233(02)00237-X

Owczarczak-Garstecka, S. C., Watkins, F., Christley, R., & Westgarth, C. (2018). Online videos indicate human and dog behaviour preceding dog bites and the context in which bites occur. Scientific Reports, 8, 7147. https://doi.org/10.1038/s41598-018-25671-7

Pedretti, G., Canori, C., Biffi, E., Marshall-Pescini, S., & Valsecchi, P. (2023). Appeasement function of displacement behaviours? Dogs' behavioural displays exhibited towards threatening and neutral humans. Animal Cognition, 26(3), 943–952. https://doi.org/10.1007/s10071-023-01742-9

Petersson, M., Uvnäs-Moberg, K., Nilsson, A., Gustafson, L.-L., Hydbring-Sandberg, E., & Handlin, L. (2017). Oxytocin and cortisol levels in dog owners and their dogs are associated with behavioral patterns: An exploratory study. Frontiers in Psychology, 8, 1796. https://doi.org/10.3389/fpsyg.2017.01796

Powell, L., Guastella, A. J., McGreevy, P., Bauman, A., Edwards, K. M., & Stamatakis, E. (2019). The physiological function of oxytocin in humans and its acute response to human–dog interactions: A review of the literature. Journal of Veterinary Behavior, 30, 25–32. https://doi.org/10.1016/j.jveb.2018.10.008

Rehn, T., Handlin, L., Uvnäs-Moberg, K., & Keeling, L. J. (2014). Dogs' endocrine and behavioural responses at reunion are affected by how the human initiates contact. Physiology & Behavior, 124, 45–53. https://doi.org/10.1016/j.physbeh.2013.10.009

Shiverdecker, M. D., Schiml, P. A., & Hennessy, M. B. (2013). Human interaction moderates plasma cortisol and behavioral responses of dogs to shelter housing. Physiology & Behavior, 109, 75–79. https://doi.org/10.1016/j.physbeh.2012.12.002

Vas, J., Topál, J., Gácsi, M., Miklósi, Á., & Csányi, V. (2005). A friend or an enemy? Dogs' reaction to an unfamiliar person showing behavioural cues of threat and friendliness at different times. Applied Animal Behaviour Science, 94(1–2), 99–115. https://doi.org/10.1016/j.applanim.2005.02.001