The M-CHAT Screening Test: What a Positive Result Actually Means
If your toddler screened positive and you are reading this tonight, here is the number you want. After the follow-up interview, roughly half of children who screen positive are diagnosed with autism. Around 95 percent have something worth acting on.
The M-CHAT Screening Test is the questionnaire most parents in the US and many other countries meet at an 18-month or 24-month check-up. Twenty questions, yes or no, filled in while you wait. It is the most widely used early autism screener in the world.
It is also routinely administered wrong, in a specific way that matters enormously to what your result means. That is most of what this page is about.
What is the M-CHAT Screening Test?
The full name of the M-CHAT Screening Test is the Modified Checklist for Autism in Toddlers, Revised, with Follow-Up. You will see it written M-CHAT-R/F. It was developed by Diana Robins, Deborah Fein and Marianne Barton, and published in its current form in 2009.
It asks a parent about everyday behaviours. Whether the child points at things to show you something interesting rather than to get it. Whether they look where you point. Whether they respond to their name. Whether they bring things over to share. Whether they look at your face when something unexpected happens.
Those behaviours were chosen because their absence in a two-year-old is among the earliest reliable signals of autism, well before language or social differences become obvious.
The instrument is copyrighted and the official version is hosted by its authors. We describe it here rather than reproducing it, because reposting the items requires a licence.
The part that gets skipped
The M-CHAT Screening Test is a two-stage instrument. That is not a detail. It is the design.
Stage one is the twenty-question checklist. Stage two is a structured follow-up interview, conducted by a professional, which goes back through only the items the parent answered in the concerning direction and asks for specific examples. Does he ever point? Can you describe the last time? What did he do next?
A great many of the apparent problems with early autism screening come down to stage two being left out.
How the M-CHAT Screening Test is scored
The M-CHAT Screening Test is scored one item at a time, 0 or 1, where 1 is the answer associated with autism risk. Most items are scored for a “no” answer, a few for a “yes”. Totals run from 0 to 20.
| Total | Risk band | What should happen |
|---|---|---|
| 0 to 2 | Low | No action needed now. Rescreen after the second birthday if the child was under 24 months. |
| 3 to 7 | Medium | Administer the follow-up interview. The result of that decides what happens next. |
| 8 to 20 | High | Skip the follow-up. Refer straight away for diagnostic evaluation and early intervention. |
In the validation sample, about 93 percent of children fell in the low band, 6 percent in the medium band and 1 percent in the high band.
After the follow-up interview, a remaining score of 2 or more is treated as a positive screen.
What a positive M-CHAT Screening Test result means
This is the section parents come here for, so here are the actual M-CHAT Screening Test numbers.
The validation study screened 16,071 toddlers at 18-month and 24-month well-child visits across metropolitan Atlanta and Connecticut, and followed up the children who screened positive.
Read those two numbers together, because separately each one misleads.
The first means a positive screen is not a diagnosis, and roughly half of flagged children will not turn out to be autistic. The second means that “not autism” very rarely means “nothing”. Language delay, global developmental delay and other concerns make up most of the remainder, and all of them respond better to early intervention than to waiting.
The instrument detected 67 cases of autism per 10,000 children screened, against 45 per 10,000 for earlier versions, while actually reducing the proportion of children referred from 9 percent to 7 percent. Better detection and fewer unnecessary referrals at the same time, which is unusual.
Why the follow-up matters so much
Without stage two, sensitivity stays roughly the same but false positives rise sharply. The authors are explicit about this.
The reason is simple once you see it. Parents answer a yes-or-no checklist quickly and under pressure, and some answers are near misses. A parent who ticks the concerning box for pointing may, when asked for an example, immediately describe their child pointing at an aeroplane last Tuesday. The follow-up catches that. The checklist alone cannot.
So if your child screened positive and nobody conducted a follow-up interview, your result is less informative than the published figures suggest. That is worth raising directly.
Who the M-CHAT Screening Test is for
The M-CHAT Screening Test is for children aged 16 to 30 months. That window is deliberate. It is also narrow.
Below 16 months, the behaviours being asked about have not reliably emerged in typical development, so the questions do not discriminate. Above 30 months, the items are too easy and a child can pass while still being autistic.
It is completed by a parent or primary caregiver. Not by a clinician observing the child, and not by the child.
What to do if your child screened positive
The authors of the M-CHAT Screening Test are unusually direct about this, and worth following.
Act now rather than waiting. Their own parent FAQ answers the question “should I wait a year and see what happens” with a flat no. Early intervention works best early, and time spent waiting is intervention not received.
Do two things at once. Request a diagnostic evaluation from a psychologist or developmental paediatrician, and separately contact your local early intervention service to ask about eligibility. Do not do these in sequence. Evaluation waiting lists are long, and in most systems a child does not need a diagnosis to start receiving support.
If your doctor is not worried but you are, act anyway. The authors say this explicitly. You can contact early intervention yourself in most places. A study of paediatric screening practice found that a positive screen often does not lead to further evaluation, which is a known gap in the system rather than a judgement about your child.
If your child screened negative and you are still concerned, act on the concern. No screener catches everything. The instrument tests for autism specifically and will miss other developmental issues entirely.
Strengths and weaknesses of the M-CHAT Screening Test
Strengths
- Validated on more than 16,000 toddlers, which is an unusually large base
- The two-stage design sharply reduces false positives without losing cases
- Free to use in clinical practice, and quick enough for a routine check-up
- Detects more cases than its predecessors while referring fewer children
- Standard in paediatric practice in many countries
- Catches developmental concerns beyond autism in around 95 percent of positives
- Could lower the average age of diagnosis substantially if used universally
Weaknesses
- The follow-up interview is frequently skipped, which is where the accuracy lives
- Useless outside the 16 to 30 month window
- Parent-report, so it depends on what a parent has noticed
- Does not screen for developmental problems other than autism
- A positive screen often fails to translate into an actual evaluation
- Performance varies across cultural and linguistic groups
- Reposting it requires a licence, so online copies are often altered or out of date
Controversies around the M-CHAT Screening Test
Whether universal screening is worthwhile
Some national bodies recommend giving every toddler an M-CHAT Screening Test. Others have concluded the evidence does not yet show that universal screening improves outcomes, as distinct from improving detection. The disagreement is genuine and it is about evidence standards rather than about whether autism matters.
The gap between screening and evaluation
An M-CHAT Screening Test is only useful if it leads somewhere. Research has repeatedly found that a substantial share of children who screen positive are never evaluated, because of waiting lists, cost, families losing contact with services, or clinicians deciding to watch and wait. A screening programme that identifies children and then loses them has achieved very little.
Equity
Accuracy varies across cultural, linguistic and socioeconomic groups. Some behaviours the instrument asks about are shaped by parenting norms that differ between cultures. Translated versions are not always formally validated in the population using them.
Online versions
Because reposting requires a licence, many copies circulating online are unauthorised, and some are altered, abbreviated or based on the older 23-item version. If you take one, use the authors’ official site. An altered M-CHAT Screening Test does not produce results the published figures apply to.
How the M-CHAT Screening Test compares to other instruments
Everything else on this site is for older children or adults. This table is for orientation rather than for choosing.
| Instrument | Age | Who answers | Purpose |
|---|---|---|---|
| M-CHAT Screening Test | 16 to 30 months | A parent | Early screening |
| ADOS-2 Test | 12 months to adult | Clinician observes | Diagnostic assessment |
| ADI-R Test | Mental age 2 plus | A parent, interviewed | Developmental history |
| RAADS-R Test | Adults | Yourself | Lifetime self-report |
| RAADS-14 Screen | Adults | Yourself | Brief self-report |
| AQ-50 Test | Adults | Yourself | Current traits |
| AQ-10 Test | Adults | Yourself | Referral triage |
| CAT-Q Test | 16 plus | Yourself | Masking |
If a positive toddler screen has made you reflect on yourself, which happens often and is one of the most common routes to adult diagnosis, the adult instruments are where to start. The RAADS-R Test asks about your whole life rather than your present. Our Neurodivergent Test covers seventeen conditions in one sitting. The Free Autism Test takes about six minutes if you want something shorter.
Frequently asked questions about the M-CHAT Screening Test
My child failed the M-CHAT. Does that mean autism?
No. After the follow-up interview, 47.5 percent of children who screen positive are diagnosed with autism. It is close to a coin flip. But 94.6 percent have some developmental concern worth acting on, so a positive result should always be followed up even though it is not a diagnosis.
What score is a fail?
There is no fail. Three to seven means the follow-up interview should be done. Eight or above means immediate referral without the follow-up. Below three means no action is indicated right now.
Nobody did a follow-up interview. Does that matter?
Yes, quite a lot. The follow-up is where most of the accuracy comes from, and without it false positives rise substantially. If your child scored in the three to seven range and nothing further happened, ask your clinician about the follow-up by name.
Can I do the M-CHAT Screening Test at home?
You can complete the checklist yourself from the authors’ official site. You cannot administer the follow-up interview to yourself, because it is a structured interview designed to be conducted by someone else. Treat a home result as a reason to talk to your doctor.
My child is 14 months. Can I use it?
Not yet. The window starts at 16 months because the behaviours it asks about have not reliably appeared before then, so the result would not mean much either way. If you have concerns now, raise them with your doctor rather than waiting for a questionnaire to become applicable.
My child is three. Is a negative result reassuring?
Not really. Above 30 months the items are too easy and an autistic child can pass. Ask for a developmental assessment instead.
My child passed but I am still worried. What do I do?
Act on the concern. The authors’ own guidance says exactly this. No screener catches everything, and the instrument only looks for autism. Contact your doctor, and contact early intervention to ask about an eligibility evaluation.
My doctor says wait and see. Should I?
The authors say no, and so does most early intervention research. Waiting costs intervention time that cannot be recovered. In most systems you can self-refer to early intervention without going through your doctor, and a child does not usually need a diagnosis to start.
How accurate is it?
Used properly, with the follow-up, it is among the best-validated early screeners available, tested on more than 16,000 children. Used without the follow-up, it over-refers considerably. Most complaints about accuracy are really complaints about incomplete administration.
Is it free?
Free for clinical use, and a practice may put it in its own electronic health record at no cost. It is copyrighted, not freeware. Distributing an electronic version, building it into software sold to others, or reposting it on a website requires a licence from the authors. Linking to the official site is explicitly allowed.
What is the difference between M-CHAT and M-CHAT-R/F?
The revised version has 20 items rather than 23, clearer wording, and the structured follow-up. It detects more cases while referring fewer children. If a site offers you 23 questions, it is using the superseded version.
What happens at the evaluation?
Usually a developmental paediatrician or psychologist, often several appointments. For a toddler the observational component is likely to be the Toddler Module of the ADOS-2 Test, and you may be interviewed about development using something like the ADI-R Test. Expect it to take time.
Should my other children be screened?
Siblings of autistic children have a higher likelihood of being autistic, and screening at the standard check-up ages is sensible. Some services screen siblings more than once. Ask your paediatrician.
Does a positive result go on my child’s record?
Screening results are recorded like any other clinical finding. In most systems this is helpful rather than harmful, because it is what opens access to support. If you have specific concerns about how records are used where you live, ask your clinician directly.
What to do next
If you are reading this after a positive M-CHAT Screening Test, the useful order is this.
Ask whether the follow-up interview was done. If not, request it.
Contact early intervention yourself, today or tomorrow, rather than waiting for a referral to work its way through. Ask for an eligibility evaluation.
Ask your doctor for a referral to a developmental paediatrician or psychologist, and ask what the waiting time is, so you know what you are planning around.
Write down what you have noticed, with specific examples and rough dates. That will be asked for at every stage, and a contemporaneous note is worth more than a memory six months old.
And then, as much as you can manage, carry on. A positive screen is information, not a prediction. Roughly half of these children do not turn out to be autistic, and of those who do, the ones identified early do better than the ones identified late. Finding out early is the good outcome here, even though tonight it will not feel like one.
Sources
- Robins DL, Casagrande K, Barton M, Chen CA, Dumont-Mathieu T, Fein D. Validation of the Modified Checklist for Autism in Toddlers, Revised with Follow-up (M-CHAT-R/F). Pediatrics, 2014;133(1):37-45. DOI · PubMed. The 16,071-child validation behind every figure on this page.
- M-CHAT official site, maintained by Diana Robins. The authorised version, scoring instructions, follow-up interview, translations and licensing terms. Use this rather than a copy.
- M-CHAT parent FAQ. The authors’ own answers on what to do after a positive screen, including their guidance against waiting.
- Robins DL, Fein D, Barton ML, Green JA. The Modified Checklist for Autism in Toddlers: an initial study investigating the early detection of autism and pervasive developmental disorders. Journal of Autism and Developmental Disorders, 2001;31(2):131-144. DOI. The original instrument.
- CDC: Screening and diagnosis of autism spectrum disorder. US guidance on screening ages and what follows a positive result.
- American Academy of Pediatrics, which recommends autism-specific screening at 18 and 24 months.
Search the literature yourself
- PubMed: M-CHAT-R/F validation and universal autism screening outcomes.
- Google Scholar: M-CHAT positive predictive value.
- NICHD: Autism spectrum disorder. US national institute overview.
- Wikipedia: Modified Checklist for Autism in Toddlers. Background and version history.