Functioning labels: why ‘high’ and ‘low’ functioning mislead
Functioning labels sort autistic people into two boxes that neither describe their needs nor predict their lives. Here is where the labels came from, what the evidence says, and how to describe support needs accurately.
This address was part of the original Identity-First Autistic campaign site (2016). The original page set out the campaign's opposition to all functioning labels, arguing that abilities vary by area and by environment, and that the labels dismiss either a person's abilities or their difficulties. The text below is a new, updated version written by the site’s current editors. How this site relates to the original campaign.
“High-functioning” and “low-functioning” are probably the most familiar shorthand used to describe autistic people, and among the most criticized. This page explains what the labels mean, where they came from, how they differ from the support levels in the current diagnostic manual, why many autistic people and a growing number of researchers want them retired, and what to say instead.
What functioning labels are
Functioning labels split autistic people into two (sometimes three) groups:
- “High-functioning” usually refers to an autistic person who speaks, has no intellectual disability, and may appear to cope with school or work. In clinical research it has typically meant an IQ score of 70 or above.
- “Low-functioning” usually refers to an autistic person who has an intellectual disability, does not rely on speech to communicate, or needs a lot of day-to-day support.
Related phrases work the same way: “mild” and “severe” autism, “mildly autistic”, “very autistic”, or “you’d never know she was autistic”.
The labels are informal. “High-functioning autism” grew up in clinical and research use and was often treated as overlapping with Asperger’s syndrome, but it has never been a diagnosis in the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders (DSM) or in the World Health Organization’s International Classification of Diseases (ICD).
What the DSM-5 uses instead: support levels
Since 2013 the DSM-5 has used a single diagnosis, autism spectrum disorder, and asks clinicians to record a severity level separately for two areas: social communication, and restricted or repetitive behaviors and interests. The levels are described in terms of support:
| Level | DSM-5 wording | Rough meaning |
|---|---|---|
| Level 1 | “Requiring support” | Without support, differences cause noticeable difficulties |
| Level 2 | “Requiring substantial support” | Marked difficulties that are apparent even with support in place |
| Level 3 | “Requiring very substantial support” | Severe difficulties that greatly affect daily life |
Clinicians also record whether autism occurs with or without intellectual impairment and with or without language impairment. The levels were unchanged in the 2022 text revision (DSM-5-TR). ICD-11, used in much of the world, takes a different route: it has no severity levels and instead specifies whether intellectual disability and functional language impairment are present.
Support levels are more specific than “high” and “low” functioning, because they describe the support a person needs rather than ranking the person. They still have real limitations:
- No agreed method. The manual describes the levels in words but gives no standard way to assign them. When Weitlauf and colleagues (2014) compared severity based on autism features with cognitive and adaptive measures in 726 children and adolescents, the different measures often disagreed.
- A snapshot, not a constant. The manual itself notes that severity can vary with context and over time, and cautions that the descriptive levels should not decide eligibility for services on their own. In practice some school systems and funders still use them that way.
- They drift back into functioning labels. “Level 1 autism” is now often used in conversation as a new name for “high-functioning”, with the same assumptions attached.
Why autistic people criticize functioning labels
The original Identity-First Autistic campaign opposed all functioning labels, and that view remains widely shared in autistic-led advocacy. The main objections are these.
Abilities are uneven, not high or low
Many autistic people have what is often called a spiky profile: strong skills in some areas alongside significant difficulties in others. A person might write fluent technical reports but be unable to make a phone call, or understand everything said to them while being unable to answer in speech. A single word cannot describe both peaks and troughs, so it ends up describing neither.
Functioning changes with the environment and over time
How well anyone functions depends on where they are and what they are asked to do. For example, an autistic software tester who works reliably in a quiet room may struggle to work at all in a loud open-plan office with flickering lights. Sleep, stress, illness, sensory overload, and life changes all matter. Autistic adults describe periods of autistic burnout, in which long-term strain leads to exhaustion and a loss of skills they previously had (Mantzalas et al., 2022). A person labeled “high-functioning” in one setting or year can need a great deal of support in another. This is why the social model of disability focuses on environments and barriers, not only on the individual.
“High-functioning” denies support
The “high-functioning” label tends to suggest that a person is only slightly affected and should be able to manage. In practice it can mean:
- support and accommodations are refused because the person “doesn’t need them”;
- difficulties are read as laziness, rudeness, or lack of effort;
- the person learns to hide their differences (often called masking or camouflaging), which many autistic adults link to exhaustion and poor mental health.
Research supports the concern. Alvares et al. (2020) studied 2,225 children and young people aged 1 to 18 at the time of their autism diagnosis. For those without intellectual disability, everyday adaptive skills (measured with the Vineland Adaptive Behaviour Scales) were well below what their IQ scores would suggest, and IQ was only a weak predictor of those skills. The authors concluded that “high functioning autism” is an inaccurate descriptor when based on IQ, and recommended abandoning the term in research and clinical practice.
“Low-functioning” denies agency
The “low-functioning” label tends to suggest that a person understands little, has little to say, and has a limited future. That can lead to:
- low expectations in education and few chances to learn;
- decisions being made about a person without asking them or offering accessible ways to answer;
- assumptions that a person who does not speak has nothing to communicate.
Many educators and advocates respond with presuming competence. The idea draws on Anne Donnellan’s criterion of the least dangerous assumption (1984): when we cannot be sure what someone understands, we should act on the assumption that would do the least harm if it turned out to be wrong. Assuming a person understands and offering them a way to communicate, such as augmentative and alternative communication (AAC), costs little if mistaken. Assuming they do not understand can silence them for years. Our guide to communication differences and accommodations covers this in more detail.
The labels rank people
Underneath both labels is an idea that some autistic people are closer to “normal”, and therefore more valuable, than others. That is a form of ableism. It can also divide the autistic community, setting people who are seen as more capable against those who are not.
What autistic people say about the terms
Surveys of autistic adults show low approval for functioning labels. In an international study of 654 English-speaking autistic adults, Keating et al. (2023) found terms such as “autism”, “autistic person”, and “is autistic” were consistently favored across countries, while there was still no single way of talking about autism that suited everyone; the authors stressed respecting individual preferences. In their guidance for researchers, Bottema-Beutel et al. (2021) describe functioning labels as stigmatizing and inaccurate, because functioning varies across areas of life, over time, and by context.
Better alternatives
Describe specific support needs
The most accurate alternative is usually to say what the person actually needs, in the area where they need it. For example: “Sam needs written instructions and a quiet place to work”, or “Amira uses a speech-generating device and needs extra time to compose answers.”
“High support needs” and “low support needs”
Some autistic people use “high support needs” and “low support needs” (sometimes shortened to HSN and LSN) to describe themselves, and many find these more respectful because they point to support rather than worth. They are still debated. Bottema-Beutel et al. list “high/low support needs” alongside functioning labels as terms that can flatten a person’s profile, and suggest describing specific strengths and needs instead. If someone uses these terms about themselves, follow their lead.
The debate over “profound autism”
In 2021 the Lancet Commission on the future of care and clinical research in autism (Lord et al., 2022) proposed “profound autism” as an administrative term for autistic people who need round-the-clock access to an adult who can care for them, typically alongside significant intellectual disability, limited language, or both. As Andrew Whitehouse summarized in The Conversation (2022), supporters argue it draws attention to a group often overlooked in research and services, while critics argue it divides autistic people into two groups using somewhat arbitrary criteria and risks reviving the problems of functioning labels. It is not an official diagnosis.
Profiles across areas of life
Some clinicians and services now describe people across several areas, such as communication, sensory needs, daily living, learning, mental health, and physical health. A profile like this can show that a person needs little support in one area and a great deal in another, which a single label cannot.
Quick wording guide
| Instead of | Try |
|---|---|
| “She’s high-functioning” | “She’s autistic” (and, where relevant, the specific support she needs) |
| “He’s low-functioning” | “He’s autistic and needs support with communication and daily living”, or “He has an intellectual disability” if that is accurate and relevant |
| “Mildly autistic” / “severely autistic” | “Autistic”, plus specific strengths and needs |
| “Nonverbal, low-functioning child” | “A nonspeaking autistic child who communicates using AAC” (or how they do communicate) |
| “Level 1, so they don’t need much help” | “Their support needs vary; in this setting they need…” |
| “You’d never know he’s autistic” | Leave it out; visible coping can hide real effort and difficulty |
For more terms to reconsider, see Beyond identity-first: ableist language and respectful alternatives. If you are unsure how a particular person wants to be described, see how to ask about language preferences.
About this page
This URL was originally part of the Identity-First Autistic campaign site, launched in 2016. The original page set out the campaign’s opposition to functioning labels: it argued that autistic people function well in some areas and less well in others, that functioning shifts with environment and accommodations, and that “low-functioning” dismisses a person’s abilities while “high-functioning” dismisses their difficulties. It called for the labels to be dropped from diagnosis and everyday description. This is a rewritten and updated version, published by the site’s current independent editors in November 2026, adding the DSM-5 support levels, later research, and current debates. It is not written by the original campaign’s authors.
Sources and further reading
- Alvares, G. A., et al. (2020). The misnomer of 'high functioning autism': Intelligence is an imprecise predictor of functional abilities at diagnosis . Autism, 24(1), 221–232.
- Weitlauf, A. S., Gotham, K. O., Vehorn, A. C., & Warren, Z. E. (2014). Brief report: DSM-5 'levels of support': A comment on discrepant conceptualizations of severity in ASD . Journal of Autism and Developmental Disorders, 44(2), 471–476.
- Bottema-Beutel, K., Kapp, S. K., Lester, J. N., Sasson, N. J., & Hand, B. N. (2021). Avoiding ableist language: Suggestions for autism researchers . Autism in Adulthood, 3(1), 18–29.
- Keating, C. T., et al. (2023). Autism-related language preferences of English-speaking individuals across the globe: A mixed methods investigation . Autism Research, 16(2), 406–428.
- Lord, C., et al. (2022). The Lancet Commission on the future of care and clinical research in autism . The Lancet, 399(10321), 271–334.
- Mantzalas, J., et al. (2022). What is autistic burnout? A thematic analysis of posts on two online platforms . Autism in Adulthood, 4(1), 52–65.
- Donnellan, A. M. (1984). The criterion of the least dangerous assumption . Behavioral Disorders, 9(2), 141–150.
Spotted an error or outdated source? Send a correction. Read our editorial standards.