28 Evidence-Based Practices for Autism

Twenty-eight focused intervention practices currently meet the criteria for evidence-based practice in autism, according to the most comprehensive review of the research to date, which synthesized nearly a thousand studies published over almost three decades.¹ That number has grown steadily as the research base has expanded, and it spans a wide range of approaches, from intensive behavioral programs to parent coaching, sensory therapies, technology-assisted tools, and adapted psychotherapy. The sheer breadth of options can feel overwhelming, and not every practice works equally well for every person or every goal.

How the List of 28 Practices Was Built

The 28 evidence-based practices come from a series of systematic reviews conducted at the University of North Carolina, each one updating the last. The most recent iteration screened over 31,000 articles and ultimately included 567 new studies, bringing the total number of analyzed studies to 972.² The prior version had identified 27 practices from a pool of roughly 29,000 articles.³ For a practice to make the cut, it needed positive results across multiple high-quality studies conducted by independent research teams. The list includes practices as varied as reinforcement, visual supports, social skills training, video modeling, exercise, and cognitive behavioral therapy. It is not a ranking; it is a catalogue of approaches with enough research backing to be called evidence-based.

What the list does not tell you is which practice is right for a specific person. The 28 practices target different skills and challenges, and many autistic individuals benefit from a combination rather than a single approach. Age matters, too. Some practices have evidence primarily for young children, others extend into adolescence and adulthood. That mismatch between “evidence-based” and “right for you” is one of the biggest sources of confusion for families navigating the system.

Early Intensive Behavioral Intervention

Early Intensive Behavioral Intervention, often called EIBI, is the most extensively studied approach for young autistic children. It is rooted in applied behavior analysis and typically involves structured, one-on-one teaching delivered for many hours per week, often starting before age five. A meta-analysis pooling results across multiple trials found large improvements in intellectual functioning and moderate improvements in adaptive behavior, with effect sizes of roughly 0.87 and 0.66 respectively.⁴ A Cochrane review found that children receiving EIBI gained about 15 IQ points more than comparison groups on average, and showed meaningful gains in language skills, though the evidence for reducing autism symptom severity specifically was weaker.⁵

An earlier meta-analysis characterized EIBI as the intervention of choice for young children with autism in the absence of other treatments with equally established efficacy, noting large effect sizes for IQ gains.⁶ The research, however, consistently rates as low-quality evidence by Cochrane standards, largely because blinding participants and families to whether they are receiving intensive therapy is practically impossible. That does not mean the results are fabricated; it means the research designs have inherent limitations that make it hard to rule out every alternative explanation. The gains are real in many cases, but they vary widely from child to child, and the IQ improvements do not always translate into equally dramatic changes in everyday functioning.

Naturalistic Developmental and Parent-Mediated Approaches

Not all behavioral interventions look like structured table-top teaching. Naturalistic Developmental Behavioral Interventions, known as NDBIs, blend developmental principles with behavioral techniques and embed learning in everyday activities like play, meals, and routines. A meta-analysis found that NDBIs produce gains in social communication, language, play, and cognition, though the effects tend to be largest for skills that are directly targeted and in contexts similar to the intervention setting.⁷ In other words, the skills do not always generalize automatically to new settings, which matters a great deal for how they play out in real life.

Parent-mediated interventions take this a step further by training caregivers to deliver therapeutic strategies themselves, rather than relying entirely on clinicians. A meta-analysis of parent-delivered interventions found small but consistent effects across domains, averaging about 0.23 on a standardized scale, with the highest effect sizes for socialization.⁸ A separate meta-analysis found a small but clinically meaningful effect on adaptive functioning as well.⁹ One of the most striking studies involved intervening with infants who had older autistic siblings, meaning they were at elevated risk. Parent coaching beginning in infancy showed a significant effect on early autism-related symptoms over a follow-up period extending to age three.¹⁰

The appeal of parent-mediated approaches is practical: they give families tools they can use throughout the day, not just during clinical appointments. They also scale better than one-on-one therapist-delivered programs. The effects are modest compared to high-intensity clinical interventions, but they add up over time and help bridge the gap between therapy sessions and daily life.

Peer-Mediated Interventions for Social Skills

Social skills training delivered through peers, rather than through adult therapists, has built a solid evidence base. A systematic review evaluating peer-mediated interventions concluded that the approach meets the threshold for “established evidence-based practice,” based on multiple strong-quality group studies conducted by independent research teams.¹¹ A randomized trial found that children receiving peer-mediated intervention showed significantly greater reductions in social difficulties compared to a control group, with particularly large effects for children with milder presentations.¹² A more recent systematic review confirmed that peer-mediated social skills interventions in school settings prove effective across studies.¹³

The logic behind peer-mediated approaches is straightforward: autistic children often struggle most with social interaction in exactly the settings where peers are present, so training those peers to be more inclusive and to facilitate interaction targets the problem where it lives. Trained peers learn to initiate play, model social behaviors, and respond positively to their autistic classmates. The effects tend to be more visible in structured environments like classrooms and less clear in unstructured situations like recess, which is an area that still needs work.

Adapted Cognitive Behavioral Therapy for Anxiety

Anxiety is one of the most common co-occurring conditions in autism, affecting a substantial share of autistic children and adults. Standard cognitive behavioral therapy works well for anxiety in the general population, but autistic individuals often need modifications: more concrete language, visual supports, reduced reliance on emotional labeling, and a highly predictable session structure. A large randomized trial compared a CBT program specifically adapted for autism against both a standard CBT protocol and treatment as usual. The adapted program outperformed both, with over 90% of participants in the adapted group achieving a positive treatment response, compared to about 81% in the standard CBT group and roughly 11% in the treatment-as-usual group.¹⁴

Adaptations that made the difference included structured scripts for social situations, early focus on physical cues of anxiety rather than abstract emotional states, and graded exposures that progressed more slowly. Even standard CBT dramatically outperformed doing nothing, but the autism-specific version added a meaningful edge, particularly for social functioning and internalizing symptoms. Case reports of late-identified autistic adolescents with social anxiety suggest this adapted approach also holds promise beyond childhood, though larger trials in adults are still needed.¹⁵

Communication Supports

For autistic individuals who are minimally speaking or nonspeaking, augmentative and alternative communication systems, often abbreviated as AAC, are a critical intervention. AAC includes everything from picture exchange systems to speech-generating devices and tablet-based communication apps. A review found that AAC was effective to highly effective for children with autism, though the field needs more rigorously designed studies with broader outcome measures.¹⁶ A persistent myth that AAC will discourage spoken language development has been contradicted by the available research; in many cases, access to AAC actually supports the emergence of speech rather than replacing it. The bigger barrier tends to be access. Families sometimes struggle to get AAC devices funded, and not all therapists are well-trained in how to implement them effectively in daily life.

Sensory Integration and Physical Activity

Sensory differences are a defining feature of autism, and Ayres Sensory Integration therapy is one of the few sensory-based approaches that has earned evidence-based status. A systematic review concluded that ASI meets the criteria for an evidence-based practice for autistic children aged 4 to 12.¹⁷ A separate review found strong evidence that ASI improves individually set goals for participation and daily functioning.¹⁸ It is worth noting that genuine ASI involves an occupational therapist guiding a child through tailored sensory-motor challenges in a specially equipped setting. Products marketed as “sensory therapy” online, such as weighted blankets and fidget tools, are not the same thing and have a much thinner evidence base.

Physical activity has also gained attention. Exercise and structured sports programs show benefits for communication, social interaction, and behavioral regulation in autistic children.¹⁹ A network meta-analysis of sleep interventions, discussed in more detail below, actually ranked physical activity as the most effective non-pharmacological strategy for improving sleep in autistic children, above both melatonin and behavioral sleep programs.²⁰ Exercise is not a targeted autism therapy in the way that, say, CBT or EIBI is, but its side-benefit profile is remarkably broad.

Sleep Problems and What Helps

Sleep disturbance affects an outsized share of autistic children, far more than their neurotypical peers. The American Academy of Neurology’s practice guideline recommends behavioral sleep strategies as first-line treatment, with melatonin offered if those strategies are insufficient and other contributing factors have been addressed.²¹ Behavioral strategies means things like consistent bedtime routines, reducing screen time before bed, and managing the sleep environment.

Melatonin is widely used and generally considered safe for this population.²² A recent network meta-analysis comparing different sleep interventions head-to-head found that daytime physical activity actually had the largest effect on sleep quality, followed by melatonin and then behavioral interventions.²³ The authors recommended a stepped-care approach: start with about 30 to 45 minutes of physical activity several days per week and parent education, then add melatonin for children who do not respond. This is a good example of how evidence-based practice does not always mean the most complex or expensive intervention; sometimes the best first step is remarkably simple.

Medication for Irritability and Challenging Behavior

No medication treats the core features of autism. The two drugs with the strongest evidence base for autistic individuals, risperidone and aripiprazole, target irritability, aggression, and self-injurious behavior. A landmark trial found that risperidone reduced irritability scores by about 57% over eight weeks, compared to about 14% with placebo, with roughly 69% of children meeting criteria for a positive response.²⁴ A systematic review and meta-analysis found risperidone and aripiprazole to be the most effective compounds for reducing irritability in autistic youth, with the largest effect sizes of any pharmacological treatment studied.²⁵ A more recent meta-analysis with formal quality-of-evidence grading rated both drugs as having high-certainty evidence for irritability reduction.²⁶

The trade-offs matter, though. Risperidone was associated with an average weight gain of about 2.7 kilograms over just eight weeks, along with increased appetite, drowsiness, and drooling.²⁷ Long-term metabolic effects are a real concern. These medications are typically reserved for situations where behavioral challenges are severe enough to interfere with safety, learning, or family stability, and they work best alongside behavioral and environmental supports rather than as standalone solutions.

Does More Therapy Always Mean Better Outcomes?

A common assumption is that more hours of intervention per week always produce better results. The evidence on this is more mixed than most people expect. A recent meta-analysis specifically examining the relationship between intervention amount and outcomes in young autistic children found no significant positive association between daily hours, cumulative hours, or total duration and the size of the treatment effect, when analyzed within intervention types.²⁸ That is a striking finding that runs counter to the common practice of prescribing 25 to 40 hours of therapy per week for young children.

Other research does find a dose-response relationship. One study examining learning objectives mastered during ABA-based treatment found that both intensity and duration were significant predictors of outcomes, with the strongest effects in academic and language domains.²⁹ The discrepancy likely comes down to what you measure and how. Hours matter for mastery of specific teaching objectives, but when you zoom out to broader developmental outcomes, the picture gets murkier. This is an active area of debate, and it has real consequences for families who may be pressured into schedules that leave little room for unstructured play, family time, or the child’s own preferences.

Telehealth Delivery

The shift to telehealth during the pandemic accelerated a trend that was already building: delivering autism interventions remotely, often by coaching parents via video. Systematic reviews have found that telehealth-delivered interventions are highly acceptable to families, comparable to face-to-face delivery in many cases, and effective for training parents and other implementers in therapeutic strategies.³⁰ Another review of telehealth-based behavior analytic interventions found favorable outcomes across all included studies.³¹ Telehealth is particularly valuable for families in rural areas or on long wait lists, and it naturally lends itself to parent-mediated models since the parent is the one doing the hands-on work while the clinician observes and guides remotely.

Cultural Responsiveness and Health Equity

Most autism interventions were developed and tested with predominantly white, English-speaking, middle-to-upper-income families. That creates a real problem when those interventions are offered to families from different cultural backgrounds without any adaptation. A meta-analysis examining culturally responsive interventions for autistic children found large, positive effects on social and communication skills as well as family outcomes like caregiver mental health and treatment fidelity.³² Interestingly, effects on general development and core autism features were not significant, suggesting that cultural adaptations may boost engagement and family wellbeing more than they change the trajectory of specific autism-related skills.

A systematic review of interventions designed or adapted for the Black and African American autism community found only eight publications meeting inclusion criteria, most of which targeted caregivers rather than children directly.³³ Researchers studying adaptations to an established parent-mediated program found that clinicians attempt to respond to family culture, psychosocial context, and specific barriers, and that these adaptations may support engagement and trust.³⁴ The research base here is thin but growing, and the takeaway is that “evidence-based” does not automatically mean “works equally well for everyone in every context.” Implementation matters as much as the intervention itself.

The Neurodiversity-Affirming Shift

Applied behavior analysis, the field underlying many of the 28 evidence-based practices, has come under sustained criticism from autistic adults. Much of that criticism centers on historical practices that prioritized making autistic behavior look more “normal,” including discouraging stimming, enforcing eye contact, and using compliance-driven techniques. A recent critical analysis highlighted the need for neurodivergent-affirming, assent-based practices that prioritize autonomy and individuality.³⁵ This is not a fringe position; it reflects a broad shift in the field toward ensuring that goals are functional and meaningful to the person receiving services rather than cosmetic. An intervention that teaches a child to tolerate a noisy cafeteria so they can eat lunch with friends is quite different from one that teaches the same child to suppress hand-flapping because adults find it unusual. The technique may overlap, but the purpose changes everything.

Dietary Interventions and the Evidence Gap

Gluten-free, casein-free diets are one of the most popular complementary approaches families try. The evidence is complicated. A meta-analysis of eight studies with about 300 participants found significant reductions in stereotypical behaviors and improvements in cognition among children on a gluten-free, casein-free diet, but no significant effects on other symptom categories.³⁶ Most included studies were single-blind, which limits how much weight you can put on the results. A separate review noted that studies demonstrating benefits are limited and that the underlying theory linking these dietary proteins to autism-related behavior has not been firmly established.³⁷ If your child has a documented gastrointestinal sensitivity to gluten or dairy, the diet makes obvious sense for that reason alone. As an autism-specific intervention, the evidence is suggestive at best and should not displace practices with stronger support.

Virtual Reality for Social Skills Training

Virtual reality is emerging as a tool for teaching social skills in safe, repeatable, and adjustable environments. A study of VR-based social skills training for children found that participants improved their social performance from baseline to the intervention phase, offering preliminary support for the approach.³⁸ Adults with autism have also rated VR-based social skills training highly for acceptability and usability, though researchers note that an adaptive, errorless approach tailored to each individual’s needs works better than generic scenarios.³⁹ VR is still in early-stage research for autism specifically. The technology is appealing because it allows someone to practice navigating a job interview or a party without the real-world consequences of getting it wrong, but we are years away from being able to say definitively how VR compares to in-person social skills groups.

Employment Support in Adulthood

The evidence base for autism interventions skews heavily toward early childhood. When it comes to employment support for autistic adults, the research is sparse but promising. An updated systematic review found that vocation-focused programs like Project SEARCH, which provides workplace internships with on-site autism-specific supports, led to higher employment rates at follow-up compared to control groups. Virtual reality job interview training also increased the number of job offers participants received.⁴⁰ An earlier review found that supported employment featuring a dedicated support worker led to higher rates of paid employment, longer working periods, and higher wages, though the methodological quality of the included studies was low.⁴¹ The gap between the volume of early childhood research and adult employment research is staggering, and it reflects a systemic tendency to invest in autism during early development while leaving adults to navigate a world that was not designed with them in mind.

1
PubMed Central. Evidence-Based Practices for Children, Youth, and Young Adults with Autism: Third Generation Review ↩︎
2
PubMed Central. Evidence-Based Practices for Children, Youth, and Young Adults with Autism: Third Generation Review ↩︎
3
PubMed. Evidence-Based Practices for Children, Youth, and Young Adults with Autism Spectrum Disorder: A Comprehensive Review ↩︎
4
PubMed. Clinically Significant Outcomes of Early Intensive Behavioral Intervention for Children With Autism Spectrum Disorders: An Individual Participant Data Meta-Analysis ↩︎
5
Cochrane Database of Systematic Reviews. Early intensive behavioral intervention (EIBI) for young children with autism spectrum disorders (ASD) ↩︎
6
PubMed. Meta-analysis of Early Intensive Behavioral Intervention for children with autism ↩︎
7
Autism Research. Understanding the Effects of Naturalistic Developmental Behavioral Interventions: A Project AIM Meta‐analysis ↩︎
8
Autism. Meta-analysis of parent-mediated interventions for young children with autism spectrum disorder ↩︎
9
PubMed Central. Parent-Mediated Interventions for Children and Adolescents With Autism Spectrum Disorders: A Systematic Review and Meta-Analysis ↩︎
10
Journal of Child Psychology and Psychiatry. Randomised trial of a parent‐mediated intervention for infants at high risk for autism: longitudinal outcomes to age 3 years ↩︎
11
PubMed Central. A systematic review of peer-mediated interventions for children with autism spectrum disorder ↩︎
12
PubMed Central. Effectiveness of peer-mediated intervention on social skills for children with autism spectrum disorder: a randomized controlled trial ↩︎
13
PubMed. Peer-Mediated Social Skills Interventions in Primary School Settings for Autistic Children in the Light of the ICF Bio-psycho-social Model. A Systematic Review ↩︎
14
PubMed Central. Cognitive Behavioral Treatments for Anxiety in Children With Autism Spectrum Disorder: A Randomized Clinical Trial ↩︎
15
Middle East Current Psychiatry. Autism-adapted cognitive behavioral therapy for social anxiety in an adolescent with late-identified autism spectrum disorder: a case report ↩︎
16
PubMed Central. The role of augmentative and alternative communication for children with autism: current status and future trends ↩︎
17
PubMed Central. A systematic review of ayres sensory integration intervention for children with autism ↩︎
18
American Journal of Occupational Therapy. Efficacy of Occupational Therapy Using Ayres Sensory Integration: A Systematic Review ↩︎
19
PubMed Central. Positive effects of physical activity in autism spectrum disorder: how influences behavior, metabolic disorder and gut microbiota ↩︎
20
Autism. Pharmacological or non-pharmacological therapies? The impact of different therapies on sleep in children with autism spectrum disorder: A systematic review and network meta-analysis ↩︎
21
PubMed Central. Practice guideline: Treatment for insomnia and disrupted sleep behavior in children and adolescents with autism spectrum disorder ↩︎
22
PubMed Central. Melatonin for children with autism spectrum disorder ↩︎
23
Autism. Pharmacological or non-pharmacological therapies? The impact of different therapies on sleep in children with autism spectrum disorder: A systematic review and network meta-analysis ↩︎
24
PubMed. Risperidone in children with autism and serious behavioral problems ↩︎
25
Pediatrics. Pharmacologic Treatment of Severe Irritability and Problem Behaviors in Autism: A Systematic Review and Meta-analysis ↩︎
26
Molecular Autism. Pharmacological and non-pharmacological interventions for irritability in autism spectrum disorder: a systematic review and meta-analysis with the GRADE assessment ↩︎
27
PubMed. Risperidone in children with autism and serious behavioral problems ↩︎
28
JAMA Pediatrics. Determining Associations Between Intervention Amount and Outcomes for Young Autistic Children: A Meta-Analysis ↩︎
29
PubMed Central. An evaluation of the effects of intensity and duration on outcomes across treatment domains for children with autism spectrum disorder ↩︎
30
PubMed Central. Using Telehealth to Provide Interventions for Children with ASD: a Systematic Review ↩︎
31
PubMed Central. Telehealth as a Model for Providing Behaviour Analytic Interventions to Individuals with Autism Spectrum Disorder: A Systematic Review ↩︎
32
PubMed Central. Examining the Efficacy of Culturally Responsive Interventions for Autistic Children and Their Families: A Meta-Analysis ↩︎
33
Autism. Autism interventions designed or adapted for the Black/African American population: A systematic review ↩︎
34
PubMed. A Mixed-methods Examination of Culturally Responsive Adaptation to an Evidence-based Parent-mediated Intervention Implemented for Autistic Children ↩︎
35
Societies. From Harm to Healing: Building the Future of ABA with Autistic Voices ↩︎
36
Nutrition Reviews. A systematic review and meta-analysis of the benefits of a gluten-free diet and/or casein-free diet for children with autism spectrum disorder ↩︎
37
PubMed Central. Gluten-Free Casein-Free Diet for Autism Spectrum Disorders: Can It Be Effective in Solving Behavioural and Gastrointestinal Problems? ↩︎
38
Journal of Special Education Technology. Virtual Reality–Based Social Skills Training for Children With Autism Spectrum Disorder ↩︎
39
PubMed Central. Virtual Reality Training of Social Skills in Adults with Autism Spectrum Disorder: An Examination of Acceptability, Usability, User Experience, Social Skills, and Executive Functions ↩︎
40
PubMed Central. Interventions for improving employment outcomes for persons with autism spectrum disorders: A systematic review update ↩︎
41
Campbell Systematic Reviews: Better Evidence for a Better World. Adult Employment Assistance Services for Persons with Autism Spectrum Disorders: Effects on Employment Outcomes ↩︎