Which ADOS-2 Module Your Child Will Get

How the clinician chooses a module

Module choice is usually made before the appointment, from the intake call or the referral paperwork, and it turns on one question: how much spontaneous language does this person use? Age enters only at the edges — separating a toddler from an older child with the same language, and a fluent child from a fluent adult. Nobody picks a module based on how autistic they expect someone to be.

The choice matters because the ADOS-2 works by placing someone in standardized social situations and watching what they do. A "press" is one of those set-ups — an activity engineered to create an opening for a social response, such as letting a wound-up toy run down and waiting, without prompting, to see whether the child looks over to share the moment. A press only works pitched at the right level. A bubble wand yields nothing useful with a fluent fourteen-year-old, and asking a two-year-old about friendship yields nothing at all.

The module can also change in the room. If the intake description does not match what the examiner sees in the first minutes, a careful evaluation switches rather than pushing on with a poor fit — a sign of a good examiner, not a mistake. All of it sits inside a larger picture: autism is identified from developmental history together with observed behavior rather than from any blood test, usually with a developmental pediatrician, a psychologist or psychiatrist, or a neurologist involved 1Ref 1Centers for Disease Control and Prevention (2024).Clinical Testing and Diagnosis for Autism Spectrum Disorder.That autism is identified from developmental history and directly observed behavior rather than a laboratory test, and that a comprehensive evaluation may involve a developmental pediatrician, a child psychologist or psychiatrist, or a neurologist..

The five modules, and who each one is for

Five modules cover the span from a barely-speaking toddler to a fluent adult. Each contains different activities and carries its own scoring algorithm, so they are not harder and easier versions of one test — they are separate instruments aimed at the same construct through age-appropriate and language-appropriate material. One administration usually runs forty minutes to an hour.

The texture differs sharply across that range. The Toddler Module and Module 1 are almost entirely play: bubbles, a toy birthday cake, a snack, free play, and moments engineered to see whether a child shares attention with someone else. Module 2 keeps toys and adds language demand. Module 3 keeps fewer materials and adds conversation, a story told from a wordless picture book, and questions about friendship. Module 4 has no toys at all. The Toddler Module is also the one that deliberately reports a range of concern rather than a classification, because behavior at that age is less stable.

What "phrase speech" and "verbally fluent" actually mean

These two terms decide the module, and both are narrower than their everyday meaning. Phrase speech means spontaneous, non-echoed utterances of roughly three words that at least sometimes contain a verb — "want more juice," not a memorized line from a show. Verbally fluent means flexible sentence-level language that can carry a conversation about something not physically present: describing yesterday, telling a story, answering an unscripted question.

• A large vocabulary is not phrase speech. What counts is spontaneous, flexible combination. A child who can label a hundred animals but does not build novel phrases has not reached it.
• A talkative child is not automatically fluent. Scripted speech, echoed phrases, and long monologues on a preferred subject are language, but not the flexible back-and-forth that Module 3 samples.
• Fluent does not mean articulate or unaffected. Many people who meet criteria on Module 3 or Module 4 have rich language. The construct being measured is social communication, not vocabulary.

Which is why the module number is not a rating, and it is worth saying so directly. A lower number means less spoken language at the time of testing — a different fact about a different domain, and one that changes as a child grows. The module is chosen to fit a person's language. It does not rank how autistic they are. A child assessed on Module 1 at three and Module 3 at seven has not moved up a level; their language changed, so the instrument that fit them changed. An outcome on the fluent modules is likewise not a milder finding, since those modules exist precisely because fluent presentations were being missed.

Where the module sits inside the whole evaluation

The ADOS-2 is one input among several, and no responsible evaluation ends with a module score alone. A full workup pairs the direct observation with a developmental and family history, caregiver report, records from school or early intervention, and usually some assessment of cognitive and language level — history and observation together rather than either standing on its own 1Ref 1Centers for Disease Control and Prevention (2024).Clinical Testing and Diagnosis for Autism Spectrum Disorder.That autism is identified from developmental history and directly observed behavior rather than a laboratory test, and that a comprehensive evaluation may involve a developmental pediatrician, a child psychologist or psychiatrist, or a neurologist..

Most families arrive through a screening pathway. Practices are asked to do general developmental screening at 9, 18, and 30 months and autism-specific screening at 18 and 24 months 2Ref 2American Academy of Pediatrics (2024).Developmental Surveillance and Screening.The screening schedule that routes families toward a diagnostic evaluation: general developmental screening at 9, 18, and 30 months and autism-specific screening at 18 and 24 months, on top of surveillance at every well-child visit., and children who flag are sent on for fuller assessment. The screeners were themselves built against the observational standard: the STAT, a play-based screener for the 24-to-36-month band, established its accuracy partly through agreement between its risk categories and ADOS classification 3Ref 3Stone WL, Coonrod EE, Turner LM, et al. (2004).Psychometric Properties of the STAT for Early Autism Screening.That the STAT is an interactive play-based Level 2 autism screener for children aged 24 to 36 months, and that its validation established agreement between its risk category and ADOS classification — showing the observational instrument functioning as the comparison standard for screeners..

Alongside the observation, most clinics send home a caregiver questionnaire. The best known is a 40-item yes-or-no screener whose items were drawn from the ADI-R interview; its validation found a cutoff of 15 most effective for separating pervasive developmental disorders from other diagnoses, while noting it does poorly at telling autism apart from related conditions 4Ref 4Berument SK, Rutter M, Lord C, Pickles A, Bailey A (1999).Autism screening questionnaire: Diagnostic validity.That the 40-item yes/no caregiver screener later published as the Social Communication Questionnaire draws its items from the ADI-R, scores in the direction of higher equals more symptomatology, was validated with a cutoff of 15 as most effective for separating pervasive developmental disorders from non-PDD diagnoses, and differentiates poorly between autism and related subtypes.. Higher scores there mean more reported symptoms, not a diagnosis. How ados-2 scoring turns an hour of observation into a comparable number is a subject of its own.

When a child speaks little or not at all

This is where families worry most, usually as a question about validity: if my child does not talk, how can anyone assess their social communication? The answer is that the Toddler Module and ADOS-2 module 1 were designed for exactly this and do not depend on speech. They sample gesture, eye contact, shared enjoyment, response to name, imitation, and play — the back-and-forth that happens before language does.

Those behaviors are the same ones the early-intervention literature treats as foundational. A landmark randomized trial in preschoolers with autism found that joint attention and symbolic play are core social-communication skills and directly targetable, with children improving in the specific skills that were worked on 5Ref 5Kasari C, Freeman S, Paparella T (2006).Joint attention and symbolic play in young children with autism: a randomized controlled intervention study.That joint attention and symbolic play are core social-communication skills in preschoolers with autism and are improvable through targeted intervention — the behaviors the toddler-level modules sample in children who are not yet speaking. A small early randomized trial.. Assessing minimally verbal children is therefore not a matter of working around missing speech; it is looking at the layer underneath speech, which is where the useful information sits anyway.

A nonspeaking evaluation is not a lesser evaluation. It samples an earlier layer of communication — the layer early intervention actually works on.

A few things help. Scheduling for the time of day when a child is most regulated matters more than parents expect, and so does telling the examiner in advance what that child's communication really looks like: which gestures they use, what a request looks like from them, what their AAC device or picture system does. Bring the device. The examiner needs to see communication in whatever form it actually takes.

Older teenagers, adults, and what to ask before the appointment

Module 4 exists because fluent older teenagers and adults were being assessed with material built for children, which is both undignified and inaccurate. It replaces play with conversation and interview-style tasks covering work or school, relationships, emotion, and daily routine, and asks the examiner to attend to reciprocity and the texture of the exchange rather than to whether toys are used correctly.

Self-report instruments often run alongside it. The RAADS-R is an 80-question adult scale spanning social relatedness, circumscribed interests, language, and sensory-motor domains, validated internationally in 779 people whose autistic participants were confirmed against diagnostic criteria together with the ADI and ADOS 6Ref 6Ritvo RA, Ritvo ER, Guthrie D, et al. (2011).The Ritvo Autism Asperger Diagnostic Scale-Revised (RAADS-R): a scale to assist the diagnosis of Autism Spectrum Disorder in adults: an international validation study.That the RAADS-R is an 80-question adult self-report scale covering social relatedness, circumscribed interests, language, and sensory-motor domains, validated in 779 people whose autistic participants met DSM criteria together with ADI and ADOS criteria. — which shows the observational instruments working as the reference point rather than as standalone answers.

A handful of ordinary questions change how the day goes:

• Which module do you plan to use, and what would change it?
• Who is in the room, and where will I be?
• What else happens that day, and how long is the whole block?
• How long until the report, and is there a feedback session?

Two realities shape the rest. Autism evaluation wait times are long enough in most regions that holding a single slot is a risk, and autism evaluation cost varies enormously by setting and by who does the testing. Both are easier to settle at the start of the autism evaluation process than halfway through it.