Why 80% of Autistic Girls Are Still Undiagnosed at 18

Women get diagnosed with autism later because the entire diagnostic system, the checklists, the screening tools, the “what autism looks like” template clinicians were trained on- was built almost entirely from research on boys. Add to that the fact that autism can present differently in women, plus a lifetime of being socialised to mask and accommodate, and you’ve got a system that was never built to see you in the first place. In the UK, girls are still being diagnosed up to six years later than boys, and roughly 80% of autistic girls remain undiagnosed by the age of 18, according to the Autistic Girls Network.

This one makes me properly frustrated, and it’s not a coincidence; it’s baked into the system.

Key takeaways

• Girls in the UK are diagnosed autistic up to six years later than boys on average, and around 80% remain undiagnosed at 18, per the Autistic Girls Network.
• The two studies that founded modern autism diagnosis, Leo Kanner’s (1943) and Hans Asperger’s (1944), were built almost entirely on male participants, and Asperger himself initially argued autism couldn’t occur in girls at all.
• Widely used screening tools, including the Autism Quotient (AQ) and the ADOS-2 were validated on samples that skewed heavily male and have shown measurable gender bias, per the Autism Research Institute.
• Masking, also called camouflaging, is a set of learned strategies (forcing eye contact, scripting conversations, copying facial expressions) that research consistently links to delayed and missed diagnosis in women.
• One Italian clinical cohort found that 75% of adults with autism received their eventual diagnosis an average of eight years after their first contact with mental health services, a gap the researchers linked partly to gender.
• A late diagnosis doesn’t mean something was wrong with you for not knowing. It means the system wasn’t designed to catch you.

If you’re reading this and recognising yourself in it, that’s exactly the kind of thing worked through in neurodivergent-affirming counselling in Swansea, and you’re welcome to book a free 20-minute consultation if you want to talk it through with me directly.

How were autism diagnostic criteria actually built?

They were built almost entirely on small groups of boys, not women, not girls, boys. Leo Kanner’s 1943 paper, the one that coined the term “early infantile autism,” described eleven children, eight boys and three girls. Hans Asperger’s 1944 study, the one so much of our early understanding of autism actually came from, focused on four boys, and he went as far as describing the condition as “an extreme variant of male intelligence,” initially concluding girls couldn’t have it at all, a view the Autism Research Institute and multiple historical reviews confirm he later revised.

So think about what that actually means. The entire blueprint for what autism “looks like,” the checklist clinicians were trained on for decades, was built almost entirely off the male experience. That’s not a small oversight. That’s the whole foundation being skewed from the start, and it’s the reason so much of what people picture when they think “autistic” is really just “autistic boy.”

Why don’t screening tools catch autism in women as easily?

Because the tools themselves were built and tested on mostly male samples, so they’re better at spotting how autism shows up in men than in women. The Autism Quotient (AQ), one of the most widely used screening measures in the UK, was validated with significantly more male participants than female ones, and research has since found it doesn’t measure identical traits across genders. The ADOS-2, often treated as a gold-standard clinical assessment, has also shown signs of gender bias, per the Autism Research Institute.

And because of that gap, there just wasn’t proper data on how autism shows up in women for decades. It’s honestly only quite recently, realistically the late 2000s onward, that research has started looking at autism in females specifically. So clinicians were working off an incomplete picture the whole time, and women were falling through the gaps because of it, not because they weren’t autistic, but because nobody had properly studied what autism in a woman actually looks like.

Does autism actually present differently in women?

Often, yes, at least compared to the male-based template the diagnostic criteria were built on, which makes it harder to spot using checklists that were never designed with women in mind. Researchers refer to this as the “female autism phenotype,” a pattern of traits and coping strategies that shows up more often, though not exclusively, in autistic women and girls. The differences aren’t usually about whether the underlying autistic traits exist, they’re about how visibly those traits get expressed, especially once masking gets layered on top.

Why does masking make it so much harder to get diagnosed?

Because masking is specifically designed, consciously or not, to hide the exact traits a clinician is trained to look for, which means well-practised masking can make a genuinely autistic woman look “not autistic enough” in a short assessment. Researchers Hull and colleagues’ 2017 study “Putting on My Best Normal” was the first to build a proper model of what’s now called camouflaging: strategies like forcing eye contact, rehearsing conversations in advance, and copying other people’s facial expressions and social scripts. Later research using the same measures found autistic adult women often score higher on camouflaging than men, though findings vary by study.

And then there’s the social layer stacked on top, because women are so often raised and conditioned to mask better, to accommodate, to smooth things over, to be the people pleaser in the room. That’s not a personality trait; that’s socialisation. Girls are taught from a painfully young age to fit in, to not make a fuss, to adapt themselves to whatever’s around them, and if you’re autistic, that conditioning can turn into a lifetime of masking so well that even you don’t clock what’s actually going on underneath it. If any of this sounds familiar, why do autistic people mask goes into the mechanics of it in more depth.

So you end up with this perfect storm: criteria built on boys, a genuine difference in how autism presents in women, and years and years of being taught to mask it on top. No wonder so many women don’t get diagnosed until their thirties, forties, sometimes later, after years of feeling like something was off and never quite being able to name it. One Italian study of adults with no language or intellectual disability found that 75% received their eventual autism diagnosis an average of eight years after their first evaluation by mental health services.

What does a late diagnosis actually change?

It doesn’t undo the years that came before it, but for most people it reframes them, turning “what’s wrong with me” into “this is why things felt so hard.” A diagnosis, formal or self-identified, can give language to a lifetime of feeling like you were working twice as hard just to keep up appearances. It’s worth being honest that a diagnosis isn’t a fix and it doesn’t erase the exhaustion of decades of masking, but it can be the start of doing things differently. If you’re in that stage right now, how it feels to be diagnosed later in life covers that transition in more detail, and support for late-diagnosed autistic women looks at what tends to actually help afterwards.

If any of this sounds like you, I want to say it plainly: there is nothing wrong with you for not knowing sooner. The system wasn’t built to see you. If you’re sitting with that right now, wondering what to do with it, that’s exactly the kind of thing worked through in counselling for autistic adults, and you’re welcome to reach out here.

Frequently asked questions

Why are so many autistic women diagnosed as adults instead of children?

Because the diagnostic criteria were built on studies of boys, the screening tools that followed were validated mostly on male samples, and decades of learned masking made autistic traits in women less visible to clinicians using that male-based template. Many women are only diagnosed after burnout, a mental health crisis, or recognising themselves in a child’s assessment.

Is autism really different in men and women, or is it just diagnosed differently?

Both. The underlying traits overlap significantly, but research into the “female autism phenotype” suggests presentation, especially masking and the type of special interests involved, often differs enough that male-based checklists miss it. It’s an active and still-developing area of research rather than a settled picture.

Can I still get help if I’m not sure I want a formal diagnosis?

Yes. A lot of people come to counselling to explore whether autism fits their experience without pursuing an official diagnosis, or while waiting on a long NHS assessment list. If you’re at that stage, I think I’m autistic, what should I do? is a good starting point.

Does masking ever fully stop once you know you’re autistic?

Not usually straight away, and often not completely, since it can be a deeply ingrained habit built up over decades. What tends to shift is your ability to notice when you’re doing it and to choose, in some spaces at least, not to.

I’m a counsellor, not a clinician, and I’m not trained or able to make any kind of diagnosis. If a formal diagnosis is something you’re seeking, I can signpost you to someone who can help with that. Nothing above is clinical advice.