How ASD Levels of Support Are Assigned and Rated

ASD levels refer to the three “levels of support” built into the DSM-5 diagnosis of autism spectrum disorder, ranging from Level 1 (“requiring support”) to Level 3 (“requiring very substantial support”). These ratings are assigned separately for two areas: social communication and restricted, repetitive behaviors. While the DSM-5 describes qualitative differences between the levels, it never provided standardized methods for deciding which level a person falls into, which has created real inconsistency in how clinicians use them.

What the Three Levels Actually Describe

The DSM-5 replaced older subcategories like autistic disorder, Asperger’s syndrome, and PDD-NOS with a single autism spectrum disorder diagnosis paired with a severity marker. That severity marker has three tiers. Level 1 describes someone who struggles with social communication but can speak in full sentences, initiate interactions, and generally manage daily life with some support. Level 2 describes more pronounced difficulties: limited speech, narrow interests that are obvious to casual observers, and noticeable distress with changes in routine. Level 3 describes individuals with very limited verbal communication, extreme difficulty coping with change, and behavior that significantly interferes with functioning across all settings.

Each level is rated independently for social communication and for restricted, repetitive behaviors. A person could, in theory, receive Level 2 for social communication and Level 1 for repetitive behaviors, though in practice many clinicians assign a single overall level. The DSM-5 itself offers qualitative descriptions of each tier but left the quantitative question open: how do you draw the line between Level 1 and Level 2?¹ That ambiguity has been a persistent issue since the manual’s 2013 publication.

How Clinicians Assign Levels

There is no single test that spits out a level number. Clinicians typically combine standardized assessment tools with their own clinical judgment. The Autism Diagnostic Observation Schedule (ADOS) is one of the most widely used instruments. Researchers have developed calibrated severity scores from ADOS data that are less influenced by a person’s age or verbal ability than raw scores, giving a more stable measure of autism-specific traits across different developmental groups.² But the ADOS was designed before the DSM-5 level system existed, so there is no direct crosswalk from an ADOS severity score to a specific support level.

Adaptive behavior scales like the Vineland-II also factor in. These measure everyday skills such as communication, self-care, and social functioning. Research in a large sample of over a thousand verbal young people with ASD found that IQ accounted for about 55% of the variation in overall adaptive skills, beyond what age and autism severity explained.³ That is a striking finding because it means how well someone functions day-to-day is driven more by cognitive ability than by their core autism symptoms. Two people with the same ADOS severity score can look very different in terms of independence, depending on their IQ.

The practical consequence is that the “level” a person receives can vary depending on which clinician does the evaluation, which tools they prioritize, and whether they weight observable symptoms more heavily than everyday functioning. The DSM-5 authors acknowledged that the levels were meant as a rough guide, not a precision instrument, but in practice they have become administrative gatekeepers: insurance approvals, school placements, and service eligibility often hinge on whether someone is classified as Level 1 or Level 2.

Why Levels Are Rated Separately for Two Domains

The decision to rate social communication and restricted/repetitive behaviors independently was intentional. Research consistently shows that these two domains do not always track together. Someone might have significant difficulty reading social cues and maintaining conversations (higher support needs in social communication) while having relatively mild restricted interests (lower support needs in that domain). The DSM-5 features these separate ratings precisely because collapsing them into a single number would obscure meaningful clinical differences.⁴

In practice, though, the two-domain system is underused. Many diagnostic reports list a single overall level, partly because service providers and insurers want one number and partly because clinicians find it cumbersome to explain split ratings to families. The result is that some of the nuance the system was designed to capture gets lost before it reaches the people who need it most.

Do Levels Change Over Time

A common question from families is whether a person’s level is permanent. The evidence says it is not. A review of the research on autism symptom severity over time found that, rather than staying stable, symptoms change in a substantial proportion of individuals. The most common pattern is decreasing severity, with most studies reporting drops in the average severity of entire groups or among large subsets of participants over time.⁵ That said, some studies identify large groups whose symptoms remain stable, and a smaller proportion of individuals show increasing severity.

This matters practically because a child diagnosed at Level 2 at age four may present quite differently at age twelve after years of intervention, maturation, and environmental adaptation. Families sometimes worry that a lower level assignment means their child will lose access to services, which can create a perverse incentive for clinicians to assign higher levels than the current presentation warrants. Some service systems are beginning to recognize that reassessment at regular intervals makes more sense than treating the initial level as a fixed characteristic.

The Gap Between IQ and Daily-Life Skills

One of the most consistent findings in autism research is that people with ASD tend to have adaptive skills well below what their IQ would predict. In other words, someone with average or above-average intelligence may still struggle with tasks like managing money, navigating public transportation, or handling routine social interactions at work. Research shows this gap widens with age: older individuals with ASD tend to have a greater mismatch between their cognitive abilities and their everyday adaptive skills.⁶

This creates a real problem for the level system. A person who tests well on cognitive assessments and speaks fluently may receive a Level 1 designation, which suggests they need only modest support. But their actual ability to hold a job, maintain relationships, and manage a household may be far more impaired than that label implies. Executive function difficulties play a role here. Adults with ASD show problems with cognitive flexibility, planning, and self-monitoring that are associated with impaired adaptive functioning and higher rates of anxiety and depression.⁷ These executive function challenges are invisible on a brief diagnostic assessment but profoundly shape daily life.

Camouflaging and Diagnostic Timing

Some autistic people learn to mask their traits in social situations, a phenomenon researchers call camouflaging. This can involve consciously mimicking neurotypical social cues, suppressing stimming behaviors, or scripting conversations in advance. Camouflaging complicates the level system because a person who is highly skilled at masking may appear less impaired during an assessment than they actually are in everyday life.

A study of 502 autistic adults found that those diagnosed in adulthood reported higher levels of camouflaging than those diagnosed in childhood, particularly on measures of assimilation and compensation. Females reported more camouflaging across all subscales compared to males, and gender-diverse adults reported elevated compensation compared to cisgender adults.⁸ The implication is that the people most likely to receive a lower support level (or to be missed entirely) are also the people expending the most effort to appear “fine,” often at significant psychological cost.

This intersects with sex and gender in important ways. Research comparing females and males with ASD across different language ability groups found that, among those with phrase or fluent speech, females tended to have lower cognitive, adaptive, and social abilities than males in the same language group.⁹ In other words, by the time girls and women are identified as autistic, they often have greater impairments than boys and men at the same verbal level, suggesting that the diagnostic process systematically misses females who present with milder or more camouflaged traits.

The “Profound Autism” Proposal

In recent years, some researchers have pushed for a fourth category beyond the three DSM-5 levels. The Lancet Commission on autism proposed the term “profound autism” for individuals with high core symptom severity, co-occurring intellectual disability, and little or no language who require extensive long-term care.¹⁰ The argument is that grouping everyone from a tech worker who struggles at parties to someone who cannot dress themselves under the same diagnostic umbrella obscures the reality that their support needs are profoundly different.

The proposal has been controversial. Supporters say it would help direct research funding and services toward those with the greatest needs, who they argue are increasingly invisible in public conversations about autism. Critics worry it would create a new binary that functions much like the old “high-functioning” and “low-functioning” split, erasing the complexity of people who fall in between and potentially reducing support for those who do not qualify for the “profound” label.

What the Autistic Community Says About Functioning Labels

The tension over levels connects to a broader debate about functioning labels. Terms like “high-functioning” and “low-functioning” have been widely used by clinicians, educators, and the media for decades. Many autistic self-advocates object to these labels, arguing that “high-functioning” is used to deny support (“You don’t look autistic, you’re fine”) while “low-functioning” is used to deny agency (“They can’t make decisions for themselves”).

Research surveying autistic people about their language preferences found that the optimal approach is not to seek consensus but to choose the option people find least disagreeable. That study concluded this means using identity-first language and not using functioning labels.¹¹ The DSM-5 levels were partly intended to replace the old functioning labels with something more descriptive and less judgmental, but in practice “Level 1” has often become a synonym for “high-functioning,” carrying the same baggage.

The Merger of Asperger’s Syndrome

The creation of the ASD level system coincided with a major and still contentious change: the elimination of Asperger’s syndrome as a separate diagnosis. Before 2013, people who had strong language skills and average or above-average IQ but significant social difficulties were often diagnosed with Asperger’s rather than autistic disorder. The DSM-5 folded Asperger’s into the broader ASD category, and the ICD-11 followed suit.

Estimates suggest that roughly 10 to 15 percent of individuals previously diagnosed with Asperger’s syndrome no longer meet ASD criteria under the DSM-5 and ICD-11, raising concerns about loss of service access and psychological wellbeing for those who fall off the spectrum under the new framework.¹² For some, the Asperger’s label was central to their identity and provided a sense of community. Its removal, even if scientifically justified by the lack of clear biological boundaries between Asperger’s and other ASD presentations, has had real social and emotional consequences.

Disparities in Who Gets Which Level

The level someone receives is not determined solely by their traits. Race, ethnicity, socioeconomic status, and geography all play a role. Research using a biosocial-ecological framework has documented cumulative influences of social injustice at both interpersonal and systemic levels on autism diagnosis.¹³ Black and Latino children in the United States tend to be diagnosed later than white children and are more likely to receive an initial misdiagnosis of a behavioral disorder rather than ASD. Later diagnosis often means a different level assignment, not because the child’s traits differ but because the assessment happens at a different developmental stage.

Socioeconomic factors compound this. Families with more resources can access private evaluations, seek second opinions, and advocate for a level assignment that matches their child’s actual needs. Families without those resources may be stuck with whatever a single overstretched clinician decides in a brief appointment. The level system, which was designed to standardize the description of support needs, can end up reflecting the family’s access to the diagnostic system as much as the individual’s actual presentation.

Sensory Processing Across the Spectrum

Sensory differences are a core feature of autism that the level system captures poorly. Some autistic people are hypersensitive to noise, light, or touch, reacting strongly to stimuli that most people barely notice. Others are hyposensitive, seeking out intense sensory input or failing to register pain and temperature changes. Research points to altered neural pathways, sensory gating dysfunction, and atypical sensory modulation as contributing to these processing differences.¹⁴

Sensory difficulties can be just as disabling as social communication challenges, yet they do not have their own level rating. A person at Level 1 for social communication might be unable to enter a grocery store because of fluorescent lighting and ambient noise. The level system, focused as it is on social and behavioral domains, misses that kind of functional impairment entirely. Some clinicians factor sensory issues into their overall level assignment, but there is no standardized way to do so.

Co-occurring Conditions and Their Influence on Functioning

Autism rarely arrives alone. Epilepsy, sleep disorders, gastrointestinal problems, and immune dysfunction all occur at elevated rates in people with ASD.¹⁵ Psychiatric co-occurrences like anxiety, depression, and ADHD are also extremely common. These conditions can dramatically increase a person’s support needs in ways the level system does not capture.

Consider someone with ASD who also has severe anxiety and chronic insomnia. Their autism-specific traits might fit neatly into Level 1, but the combined burden of all three conditions could leave them unable to work or live independently. The level assigned by a clinician focused on autism criteria alone does not reflect that reality. This is one of the strongest critiques of the level system: it describes autism in isolation, while the person lives with everything at once.

What Levels Mean for Adult Outcomes

Research on young adults with autism paints a nuanced picture of how levels relate to outcomes in employment, social life, and wellbeing. A longitudinal study comparing autistic adults with higher cognitive abilities to those with lower cognitive abilities found that verbal intelligence was related to work outcomes, while adaptive skills and friendships were more closely linked to autistic features. Across both groups, adults with ASD tended to report lower wellbeing and fewer positive emotions compared to non-autistic peers.¹⁶

The picture for employment is especially challenging. Workplace accommodations that autistic employees find helpful include adjustments to the interview process, flexible scheduling, quiet workspaces, and training for coworkers about autism.¹⁷ Many of these accommodations are inexpensive and straightforward, but accessing them requires disclosure, which many autistic adults are reluctant to do. The level system is largely irrelevant in this context. An employer is not going to look at a Level 1 or Level 2 designation and know what accommodations to provide. What matters is a conversation about the specific person’s specific needs.

Genetic Architecture and Severity

Research has begun to uncover genetic patterns that map loosely onto the severity spectrum. A study of more than 2,000 simplex ASD cases (families where only one child is affected) found that lower IQ was associated with a higher rate of spontaneous gene mutations, while higher IQ was associated with a stronger family history of psychiatric conditions like depression, bipolar disorder, and schizophrenia. Individuals who could not complete a traditional IQ test, a group with the most severe language and behavioral impairments, had the highest rates of new mutations and the lowest rates of family psychiatric history.¹⁸

The implication is that autism at the most severe end of the spectrum may be driven more by rare, disruptive genetic events, while autism at the milder end may arise from a larger number of common genetic variants shared with other psychiatric conditions. This does not mean the level system maps cleanly onto genetic categories. Biology is far too messy for that. But it does suggest that what we call “ASD Level 1” and “ASD Level 3” may represent meaningfully different biological pathways, not just different points on a single continuum.

Caregiver Burden Across Support Levels

The impact of a child’s support level on family life is substantial and unevenly distributed. A study of caregivers of children with ASD found that about 42 percent experienced mild to moderate burden, roughly 34 percent experienced high burden, and only about 25 percent reported no to mild burden. The burden fell disproportionately on female caregivers: over 41 percent of female participants reported high burden compared to about 14 percent of males.¹⁹

Higher support levels generally mean greater caregiver burden, but the relationship is not as linear as you might assume. Families of children at Level 1 can face intense stress from behavioral challenges, social isolation, and the exhausting work of navigating school systems and insurance. The difference is that their challenges are less visible, which often translates to less sympathy and less help from extended family and community.

Communication Support Beyond Speech

For individuals at Level 2 and Level 3, augmentative and alternative communication (AAC) tools can be transformative. These include picture exchange systems, speech-generating devices, and apps that allow someone to communicate by selecting symbols or typing. Children with the most significant learning needs stand to gain the most from AAC interventions, yet comparisons between AAC and speech-only approaches must account for the fact that these children also pose the greatest challenges to achieving large and consistent effects from any intervention.²⁰

A persistent myth holds that providing AAC tools will discourage a child from developing speech. Research has not supported this concern. In many cases, AAC acts as a bridge that reduces frustration and behavioral challenges while spoken language continues to develop. For families navigating the level system, pushing for AAC access should not depend on a Level 3 designation. Children at Level 2 and even some at Level 1 who have significant expressive language difficulties can benefit, but getting that written into a support plan often requires advocacy that many families find overwhelming.