Autism is detected through a combination of observing specific behavioral patterns and completing a formal evaluation with a qualified professional. There is no blood test or brain scan that diagnoses autism. Instead, detection relies on recognizing early signs, completing standardized screenings at recommended ages, and pursuing a comprehensive clinical assessment when concerns arise.
Early Signs in Infants and Toddlers
Many signs of autism become visible before a child turns two, though they can be subtle and easy to dismiss. The earliest red flags involve social communication: not responding to their name by 9 months, not showing facial expressions like happiness or surprise by 9 months, and not using gestures like waving goodbye by 12 months. By 15 months, most children will hold up objects to share their excitement with you. A child who doesn’t do this, or who doesn’t point at something interesting by 18 months, may be showing early signs worth tracking.
By 24 months, children typically notice when someone around them is hurt or upset. A lack of this awareness can be another indicator. Alongside these social signs, you may notice repetitive behaviors: lining up toys and getting distressed when the order changes, repeating the same words or phrases, fixating on parts of objects (like spinning the wheels on a car rather than playing with it), or flapping hands and rocking the body. Unusual reactions to sounds, textures, smells, or lights also fall into this category. No single behavior on its own means a child is autistic, but a pattern across several of these areas is a strong signal to pursue screening.
Recommended Screening Ages
The American Academy of Pediatrics recommends that all children be screened for developmental delays at their regular well-child visits at 9, 18, and 30 months. On top of that, every child should be screened specifically for autism at 18 months and again at 24 months. These screenings happen regardless of whether a parent has concerns, because some signs are easier for a trained provider to spot than for a parent who sees their child every day.
The most widely used screening tool for toddlers is the M-CHAT-R, a short questionnaire validated for children between 16 and 30 months (sometimes used up to 48 months). It asks parents about their child’s behavior in everyday situations. A positive result on this screener doesn’t mean a child is autistic. It means a more thorough evaluation is warranted. Screening is the first filter, not the final answer.
What a Formal Evaluation Looks Like
A formal autism evaluation is conducted by specialists such as developmental pediatricians, child psychologists, or neuropsychologists. The process typically involves a parent interview, a review of the child’s developmental history, and direct observation using standardized tools.
The gold-standard observational tool is a structured assessment where a clinician introduces specific activities designed to draw out social interaction, communication, and repetitive behaviors. For young children or those with limited language, this might include playing with bubbles, releasing an inflated balloon, or setting up a pretend birthday party. For older children, adolescents, and adults with more developed language, the assessment shifts to conversations about emotions and relationships, retelling a story from a book, or demonstrating a daily routine. The clinician watches how the person engages, responds, initiates interaction, and handles transitions between activities.
To receive an autism diagnosis, a person must show persistent difficulties in all three areas of social communication: back-and-forth social interaction, nonverbal communication (eye contact, gestures, facial expressions), and developing and maintaining relationships. They must also show at least two of four types of repetitive or restricted patterns: repetitive movements or speech, rigid adherence to routines, intensely focused interests, or heightened or reduced sensitivity to sensory input. These patterns need to have been present from early development, even if they weren’t recognized at the time.
How Autism Presents Differently in Girls
Girls and women with autism are frequently diagnosed later than boys, or missed entirely. One key reason is that their restricted interests often look less obviously unusual. Boys are more likely to develop intense fixations on vehicles, screens, or mechanical systems, which stand out to parents and teachers. Girls tend to develop intense interests in people, animals, or craft activities, interests that appear typical on the surface even when the intensity behind them is not.
Girls are also more likely to engage in social camouflaging, actively concealing traits that peers might view as unusual. This can include carefully observing and copying how other people behave in social situations, scripting conversations in advance, and forcing themselves to maintain eye contact even when it feels uncomfortable. This masking can be convincing enough to influence what a clinician observes during an assessment, creating a gap between what parents report at home and what the evaluator sees in the office. If you’re a parent of a girl who seems to struggle socially at home but “holds it together” at school or in structured settings, this discrepancy itself is worth raising with a provider.
Detecting Autism in Adults
Many adults discover they may be autistic after years of feeling different without understanding why. This is especially common among people who developed effective masking strategies early in life. Adult camouflaging typically involves three overlapping behaviors: compensating for social difficulties by using rehearsed scripts and copying others, masking autistic traits by constantly monitoring your own eye contact, facial expressions, and gestures to present a neurotypical appearance, and assimilating by forcing yourself to interact even when it feels artificial.
These strategies can work well enough to avoid detection in childhood, but they carry a cost. Short-term camouflaging causes significant exhaustion and anxiety. Over years, it affects mental health, self-perception, and even access to appropriate support, because the person appears to be functioning without difficulty. Adults who recognize themselves in these descriptions often report that social interaction feels like performing a role rather than being themselves. Many are initially misdiagnosed with anxiety disorders or depression, which may be real but secondary to the underlying autism.
Adult evaluation follows the same diagnostic framework as childhood assessment, but the clinician relies more heavily on self-reported history, asking about childhood experiences, sensory sensitivities, social patterns, and how you navigate daily routines. Bringing a parent or someone who knew you as a child can strengthen the evaluation, since the diagnostic criteria require that traits were present in early development.
Conditions That Look Similar
ADHD and autism share enough surface features that they’re frequently confused, and they also commonly co-occur. Both can involve difficulty with social cues, sensory sensitivities, and trouble with transitions. The distinguishing factor is the underlying pattern. Autism centers on differences in social communication and restricted, repetitive behaviors or interests. ADHD centers on inattention, impulsivity, and hyperactivity. A child with ADHD might miss social cues because they weren’t paying attention. A child with autism might miss them because they process social information differently.
Sensory processing issues appear in both conditions but follow different trajectories. Children with autism tend to show some decrease in sensory sensitivities as they grow older, while children with ADHD tend to show increasing sensory issues over childhood, particularly with auditory processing. These patterns can help clinicians tease apart the two, though a thorough evaluation is the only reliable way to distinguish them.
Eye-Tracking Technology as a Detection Tool
Researchers are studying whether eye-tracking technology could eventually supplement traditional autism screening. The idea is straightforward: track where a child looks when shown social scenes or geometric patterns, and use those gaze patterns to identify differences associated with autism. Some studies have achieved strong individual results, with classification accuracy around 78 to 89 percent in controlled settings.
The technology is particularly good at confirming that a child is not autistic, with a pooled specificity of 92.3 percent across studies. But its ability to correctly identify children who are autistic remains low, with pooled sensitivity at just 35.9 percent, meaning it would miss roughly two out of every three autistic children. For now, eye-tracking is a promising research tool but not a replacement for clinical evaluation. It may eventually serve as one piece of a broader screening process, particularly in primary care settings where access to specialists is limited.