Can you be autistic and still make eye contact?

Eye contact also isn’t a diagnostic criterion. DSM-5-TR mentions abnormalities in eye contact as one example of poorly integrated verbal and nonverbal communication, under a criterion about social communication as a whole. Nothing in the criteria requires a particular gaze pattern.

It isn’t specific to autism either. Reduced eye contact is documented in social anxiety disorder, and a 2025 analysis notes that qualitative differences in eye gaze can appear in ADHD, anxiety, psychosis, or depression. Eye contact is a weak signal in both directions, which is why assessments don’t rest on it.

What the criteria say, and what they don’t

The belief that autistic people cannot make eye contact can lead some adults to rule themselves out before an assessment. The diagnostic criteria give eye contact a narrower role than that stereotype suggests.

DSM-5-TR’s social communication criterion requires three features, all present: deficits in social-emotional reciprocity, deficits in nonverbal communicative behaviours used for social interaction, and deficits in developing, maintaining, and understanding relationships. Abnormalities in eye contact appear as an illustrative example under the second of those, alongside body language and difficulties understanding and using gestures (American Psychiatric Association, 2022).

An example is not a requirement. The criterion is about the integration of verbal and nonverbal communication, and eye contact is one of several ways that difference might show. Someone can meet that criterion through gesture, facial expression, prosody, or the coordination between them, with eye contact playing no part in it.

Reduced eye contact is discussed in early-development research and screening, but a group-level signal in young children is not an adult diagnostic rule. DSM-5-TR also states that traits may be masked by learned strategies later in life. Neither point makes eye contact decisive in either direction.

Eye contact varies more than the stereotype allows

The evidence that autistic people don’t uniformly avoid eye contact has been building for some time, and one recent study is worth describing in detail because of how it was designed.

Researchers used a live face-to-face setup instead of screen-based stimuli, combining a wearable eye-tracker with electrodermal activity sensors to measure both gaze and physiological stress during a real conversation. The autistic participants didn’t differ from the non-autistic group in their eye behaviour or in their skin conductance responses. What the study did find was that non-autistic participants appeared considerably more distressed than autistic participants when their conversation partner didn’t look at them (Clin and Kissine, 2023).

That finding cuts two ways worth naming. It undermines the assumption that autistic people can be identified by gaze in a natural conversation. It also relocates part of the discomfort around eye contact to the non-autistic side of the interaction, which the authors argue means social interaction difficulties should be treated as relational, not as an individual deficit.

This is one study with a specific design, and it doesn’t overturn the broader literature reporting group differences in gaze. Participants defined words in a highly predictable setting, their attention was drawn to printed cards, and the authors note that a floor effect may have hidden subtle differences. The finding is limited to this task; it does not establish the reliability of gaze differences in ordinary conversation or assessment.

What autistic people describe about eye contact

Where the experience is difficult, first-hand accounts describe something more specific than dislike.

A qualitative analysis of self-reported accounts from teens and adults found several recurring themes: adverse emotional and physiological reactions, feelings of being invaded, and sensory overload during eye contact, alongside difficulty understanding social nuance and difficulty both receiving and sending nonverbal information while holding someone’s gaze (Trevisan et al., 2017).

The same analysis reported something that matters more for this question. Participants described strategies they had devised to cope with or work around eye contact difficulties. People had built their own methods, which is the mechanism by which someone can find eye contact aversive and make it anyway.

Two limits belong with that study. Participants had self-declared autism, not confirmed diagnoses, and qualitative accounts describe experience without establishing how common each pattern is. What it does well is show the range, which is the opposite of a uniform picture.

The accounts also contained strategies for managing eye contact, including exposure and practice, counting, looking near rather than at the eyes, verbal or nonverbal backchanneling, and using eye contact strategically. Because the study analysed public posts and videos from people with self-declared autism, it cannot establish how common these strategies are or when people learned them.

Camouflaging research covers the same general distinction from another direction. Researchers measure both observable presentation and self-reported effort (Cook et al., 2021). The review does not rank eye contact as the most common strategy, but the distinction still matters: an observed behaviour and the effort or method behind it are different sources of information.

Why eye contact is a weak signal either way

Making eye contact doesn’t rule autism out. Avoiding it doesn’t rule autism in.

A systematic review and meta-analysis of eye-tracking studies in social anxiety disorder identified 30 articles and found that patients made significantly less eye contact with negative emotional faces than healthy controls, with a moderate effect, and significantly less with positive emotional faces as well (Günther et al., 2021). Reduced eye contact is a well-documented feature of social anxiety.

A 2025 analysis in the British Journal of Psychiatry makes the general point while discussing how impairment gets interpreted in autism diagnosis, noting that qualitative differences in eye gaze could be present in the context of ADHD, anxiety, psychosis, or depression (Hollingdale, Woodhouse and Deeley, 2025). That’s a short analysis piece, not a study producing a prevalence estimate.

The same article notes that some people may camouflage autistic traits enough that apparent impairment falls below a diagnostic threshold despite the effort involved. It does not test eye contact specifically or rank its diagnostic reliability, so those broader claims should not be drawn from the article.

What to do with this if you’re weighing an assessment

Five things follow.

Separate the two questions the stereotype fuses. Whether you make eye contact and whether it is effortless are different and potentially informative questions. Neither answers the diagnostic question alone.

Don’t use eye contact as a self-screening test. It’s the wrong instrument in both directions, and using it is how people talk themselves out of a question worth asking. The criteria are about social communication as a whole, across contexts, from early development, alongside a second required domain of restricted and repetitive behaviour that has nothing to do with gaze.

Describe the behaviour and its context. What you do is one source of information; any effort, discomfort, strategy, or change across settings adds context a clinician may not be able to observe. Counting before looking away, looking near the eyes, or finding gaze harder under particular demands are all reportable. Our page on whether an assessment can miss you if you mask covers how to raise that.

Do not let one observation become a verdict. Making eye contact in one conversation cannot rule autism out. Observation of nonverbal communication may contribute to a comprehensive assessment, but it has to be interpreted with developmental history, context, other criteria, and alternative explanations.

Expect the question to come up, and answer it plainly. Assessments do look at nonverbal communication, and eye contact is part of what gets observed. What helps is context: what you do, what it takes, and what changes when you’re tired, in an unfamiliar setting, or with someone you know well.

For the wider picture of what masking is and what it costs, our guide to autistic masking goes further. If you want to know what the appointment itself involves, what happens in an adult autism assessment covers it, and the NeuroDiversion AuDHD self-reflection is a structured way to gather your own examples first. Neither is a diagnosis, and nothing here replaces a clinician.