She was fine at school. The teacher said so, warmly, twice. Then she walked through the front door, dropped her bag, and screamed for forty minutes about a sock.
If you are reading about signs of autism in girls, there is a decent chance you have already been told she is fine: by a teacher, a relative, or a paediatrician who watched her sit politely on a chair for eleven minutes. You are the only person seeing the other half of the day, and the gap between those two halves is not you being dramatic. It is the single most diagnostically important thing about your daughter, and nobody in the room has the vocabulary for it yet.
This piece is about that gap: what it looks like, why the system misses it, and how to put it in front of a clinician in a form they can use.
Why is autism missed in girls?
Because the diagnostic picture was built mostly from boys, and because many autistic girls compensate well enough to slip under the threshold that picture describes. The behaviours clinicians were trained to look for are the ones that disrupt a classroom. A girl who copies the popular child, memorises the rules of conversation, and holds herself together until 3:30pm disrupts nothing.
There is a second reason, less often said: the compensating itself is read as evidence against autism. She made eye contact, so no. She has a best friend, so no. She is chatty, so no. Each of those observations is true and none of them rules anything out.
Two honest caveats. On the girls versus boys question, this pattern is not exclusive to girls: boys mask too, and framing it as a girl thing has caused its own missed diagnoses. And the research on sex differences in autism presentation is younger and far more contested than most articles let on. Clinicians widely observe the pattern; the field is still arguing about how much is biological and how much is social expectation. Treat anyone who says it is settled with mild suspicion.
What masking actually looks like
Masking is not deception and it is not a choice she made. It is a set of learned strategies for surviving environments that do not fit, mostly below conscious awareness.
Around age 5, masking is imitative and clumsy. She copies another child's play almost move for move, a beat behind. She hovers at the edge of a group and joins by mirroring rather than by initiating. Adults describe her as shy, sensitive, or a bit of a follower.
By nine it is deliberate, and expensive. She has studied how girls talk to each other the way other children study for a test: a rehearsed set of moves, a mental list of rules, do not talk about the horses for more than two minutes, laugh when they laugh. Friendships shift from parallel activity to alliances, and what worked at five stops working. This is when anxiety, stomach aches, school refusal and perfectionism arrive, and why so many autistic girls collect an anxiety diagnosis years before anyone says autism.
Signs of autism in girls that look different from the textbook list
Three more that get filed under something else.
Sensory distress reads as fussiness: seams, labels, hair brushing, food textures, the smell of the dining hall. Imaginative play is rich but rule-bound, replayed the same way, with resistance when another child deviates. And a precise, adult-sounding register learned from books rather than from peers gets written up as advanced verbal skill, which is where autism and speech delay earns a read: a large vocabulary and strong communication are not the same thing.
How to document masking for a 90-minute appointment
A clinician sees her in a quiet, novel, one-to-one setting with an adult who follows her lead and asks interesting questions. Close to optimal conditions for an autistic girl. She will very likely perform beautifully. Your job is not to argue with what they saw. It is to supply the data they cannot collect.
Start two months out if you can. If the appointment is next week, start now anyway.
1. Keep a dated log, not a diary — one line per entry. Date, time, place, what happened, how long it lasted. "14 Sept, 3:40pm, hallway, screaming after school, 35 min, triggered by asking about her day." Twenty entries of that beat five pages of narrative.
2. Film the contrast — a short clip of her at ease at home doing her thing, and, if you can do it without making her self-conscious, a clip of the first two minutes after she gets home: the bag drop, the shoes, the moment before it turns. The build-up is more useful to a clinician than the peak, and easier to film. Offer the clips rather than waiting to be asked.
3. Count things — how many days this month she came home and needed to be alone before speaking. How many mornings her stomach hurt. How many playdates last term, and who initiated them. Numbers you actually counted are unassailable in a way that "often" is not.
4. Get the school's version in writing — email the teacher specific questions, not general ones. "Does she initiate conversation with peers, or respond when approached?" "Who does she play with at break?" General questions get "she's lovely". Specific questions get evidence.
5. Write the one-page summary — three sentences on why you are here, then your top six observations with an example and a date for each, then the sentence that matters: "She presents very differently in a one-to-one setting with an interested adult, and I would like that considered." Hand it over at the start of the appointment, not the end.
6. Ask directly whether the clinician assesses masking — someone who works with high-masking presentations will have a ready answer. Someone who does not will tell you she seems very social. Ask for a speech-language evaluation alongside the autism assessment too: social communication sits squarely in an SLP's scope, and ASHA's autism practice portal sets out what that covers.
What to say instead: a phrase table for the appointment
The words you choose shape whether you are heard as a reporter or as an anxious parent. Unjust, and also true.
| Instead of | Say |
|---|---|
| She's really struggling | Here are eleven dated examples from the last six weeks |
| She has meltdowns | She has extended distress episodes, twenty to forty minutes, almost always within an hour of getting home |
| She's very sensitive | She removes labels from clothing, cannot tolerate hair brushing, and eats four foods reliably |
| She's obsessed with horses | Horses account for most of her conversation, her play, and her free time, and interrupting causes real distress |
| I think it might be autism | I am asking for an autism assessment, and here is what I have observed |
| The school says she's fine | The school sees her regulated. I see what it costs her. Here is what that looks like at 4pm |
The cost of leaving it
There is a version of this where you decide it does not matter, because she is coping. Worth naming rather than dismissing — for some families it is the right call, and a diagnosis brings little they do not already have.
But the cost of long-term masking is reported consistently by autistic women looking back: burnout, anxiety, depression, a self-concept built around being defective at something other people find effortless. A child told she is fine while she is drowning learns that the drowning is a personal failing. Identification does not make her autistic. It makes her autistic and explained, which is a different childhood.
Do not confuse identification with a plan to change her. The goal is not to teach her better masking with a professional's help. It is to reduce what she has to mask: adjust the environment, protect recovery time, drop the demands that produce nothing but compliance, and let her have the horses.
Late diagnosis, and girls who are already teenagers
Plenty of high-masking autistic girls are not identified until secondary school, and plenty not until adulthood, frequently after their own child is assessed. Social demands step up, the strategies stop scaling, and the cost surfaces as anxiety or school refusal rather than as autism.
Nothing here expires at eleven. The documentation system above works as well at fourteen as at five, and a teenager is often the better witness to her own masking. Ask her what a school day costs her, then be quiet for longer than is comfortable. She may have been waiting a while for someone to ask.
FAQ
Can autism be diagnosed if she does well at school?
Yes. Academic performance is not part of the diagnostic criteria. Doing well at school while falling apart at home is a recognised presentation, though it takes a clinician who is looking for it.
What if the assessment says she is not autistic?
One session with one clinician does not settle the question for a high-masking child, particularly if the session went well. Ask what weighting was given to your home log and to the school's written answers, and ask for that in writing. If it comes back thin, you have grounds to request a reassessment at the next step up in social demands. That is a reasonable ask, not doctor-shopping.
Is masking bad? Should I tell her to stop?
Masking is a survival skill, and telling a child to drop it in an environment that punishes difference is not fair to her. What helps is reducing the need for it in the places you control, and being explicit that she does not have to perform for you.