Social Skills Training for Autistic Kids: What Research Shows

Social skills training is one of the most recommended and most delivered interventions in the autism support landscape. Hundreds of programs operate across schools, outpatient clinics, and community centers, teaching autistic children how to make eye contact, read facial expressions, take conversational turns, and navigate peer interactions. The market for these programs is substantial, and the demand from parents who watch their children struggle to make friends is entirely understandable.

The research on whether these programs work is real, meaningful, and more complicated than the phrase “social skills training” suggests. The complications involve not just the evidence but also a set of questions the field is actively grappling with: what counts as social success, whose definition of social competence we’re using, and whether teaching autistic children to mask their natural social style produces wellbeing or harm. Parents trying to make good decisions for their children deserve an honest account of all of it.

Key Takeaways

• The PEERS program (Program for the Education and Enrichment of Relational Skills) has the strongest randomized trial evidence base of any social skills training program for autistic youth.
• Gains in social knowledge and observed behavior in structured settings are consistently demonstrated; generalization to naturalistic peer settings is less reliable and varies by child.
• Adolescents and young adults show stronger and more durable gains than younger children in most PEERS research, partly because they have greater cognitive capacity to apply learned strategies.
• A significant body of autistic self-advocate critique argues that programs focused on mimicking neurotypical social behavior teach masking — a skill with documented mental health costs — rather than authentic connection.
• Program quality varies enormously; programs using behavioral rehearsal, parent involvement, and naturalistic generalization practice produce stronger outcomes than those relying on instruction alone.
• The research is strongest for social knowledge, moderate for in-session behavior, and weakest for real-world friendship quality and frequency.

What PEERS Research Actually Shows

The Program for the Education and Enrichment of Relational Skills, developed by Elizabeth Laugeson and colleagues at UCLA, is the most extensively studied manualized social skills program for autistic youth. PEERS is a 16-week program delivered to youth in group sessions with parallel parent groups — the parent involvement is a key component, designed to help families support practice and generalization outside sessions.

Laugeson and colleagues published the foundational randomized controlled trial in 2012 in the Journal of Autism and Developmental Disorders. The study enrolled 28 adolescents with autism spectrum disorder (ages 12–17) who were randomly assigned to PEERS or a waitlist control. Treatment group participants showed significantly greater gains on the Test of Adolescent Social Skills Knowledge (TASSK), social responsiveness measures, and parent-reported quality of friendships at post-treatment. Effect sizes were large (Cohen’s d = 0.81–1.48 on primary outcomes).

A follow-up study by Laugeson et al. (2014) examined 1–5 year outcomes for PEERS graduates and found that gains in social knowledge were maintained, though the friendship quality improvements were more variable. Participants who had continued social opportunities — school environments where they had access to peers — showed more durable gains than those with fewer naturalistic social contexts.

A 2017 multicenter RCT by Schohl and colleagues examined PEERS in community clinic settings with 88 adolescents with ASD. Results replicated the original trial’s social knowledge gains and added behavioral observation data showing increased social engagement. Importantly, this study was conducted in real-world clinic conditions, not a university research lab, providing some external validity support.

The Generalization Problem

The most consistent limitation in the social skills training literature — not just PEERS but across programs — is the generalization problem. Gains that are robust in structured, therapeutic settings frequently fail to transfer to naturalistic peer environments: school hallways, lunch tables, neighborhood play, sports teams.

Researchers categorize this as the difference between acquisition (learning a skill) and performance (applying it in the right context). A child can demonstrate correct conversational turn-taking in a role-play exercise with a therapist and then revert to their natural patterns in an unstructured setting with peers. This is not willful non-compliance; it reflects the significant cognitive load of applying learned social scripts to the unpredictable, fast-paced reality of peer interaction, where many signals are processed simultaneously and timing matters in milliseconds.

White and colleagues (2007) reviewed 14 social skills intervention studies for youth with autism and found that while all studies reported positive outcomes on at least one measure, only a minority included generalization measures assessing real-world social behavior, and those that did consistently found smaller effects than in-session measures.

The factors that predict better generalization include: parent and sibling involvement in practicing skills at home, naturalistic social opportunities built into the program design (not just role-play), longer programs with more practice repetitions, and child characteristics including higher baseline cognitive ability and language level.

This is why program components matter as much as program names. A PEERS-branded program that delivers only the youth group component without the parent group, without homework practice assignments, and without naturalistic social opportunities will predictably show weaker generalization than the full protocol.

Program Feature	Effect on Generalization	Evidence Base
Parent group participation	Strongly positive	Laugeson et al. 2012, 2014
Homework practice between sessions	Positive	Multiple PEERS trials
Naturalistic social opportunities (real peer interaction in program)	Positive	White et al. 2007 review
Behavioral rehearsal (role-play) alone	Limited	Gresham et al. 2001
Instruction only, no practice	Minimal	Gresham et al. 2001
Follow-up booster sessions	Positive	Laugeson et al. 2014
Youth cognitive/language level	Moderating variable	Rao et al. 2008

The Masking Critique: A Serious Consideration

A significant and growing body of autistic self-advocate scholarship argues that social skills training — particularly programs focused on teaching eye contact, minimizing stereotypies (self-stimulatory behaviors), and mimicking neurotypical interaction patterns — produces masking rather than authentic social development.

Masking, also called camouflaging, refers to the effortful suppression of autistic social behaviors to present as neurotypical. It is cognitively and emotionally costly. Cassidy and colleagues (2018), in a study of 111 autistic adults, found that greater camouflaging was associated with significantly higher rates of depression, anxiety, and suicidal ideation, independent of autism severity. A 2020 study by Lai and colleagues in Autism found masking associated with poorer mental health outcomes in a large sample of autistic adults.

The critique is not that autistic children should not learn social skills. It is that there is a difference between expanding a child’s genuine social repertoire — helping them find authentic ways to connect that align with their own social style — and teaching performance of neurotypical norms in a way that produces exhaustion, inauthenticity, and long-term mental health costs.

Laugeson and other PEERS researchers have engaged with this critique. Their argument is that social knowledge does not require masking: knowing when to make eye contact is not the same as forcing constant eye contact, and understanding turn-taking conventions is not the same as suppressing one’s personality. They argue the program builds a toolkit, and competent delivery respects autistic individual variation in how skills are applied.

The critique is most valid when programs prioritize appearance over function — teaching a child to look like they’re listening rather than teaching strategies for genuinely engaging. Parents evaluating a program should ask directly: does this program teach kids to mask, or does it help them build authentic social tools? The distinction is visible in how programs address stimming (self-stimulatory behaviors), directness, and special-interest conversation.

Children who are experiencing social difficulties alongside other challenges — dual diagnosis, for instance — may be navigating a particularly complex picture. The AUDHD research shows that co-occurring ADHD and autism creates a distinctive social profile that benefits from interventions designed with both presentations in mind.

What Age Groups Benefit Most

The research is clearest for adolescents (ages 12–17) and young adults (ages 18–23). Several factors explain this: older youth have sufficient metacognitive capacity to understand and apply explicit social rules, they are in developmental stages where peer relationships are particularly salient, and they can articulate their own goals for social connection — making them more actively engaged participants.

Younger children (ages 5–11) show gains on some measures, particularly behavioral imitation and role-play performance, but generalization data is weaker and the interventions most effective at younger ages tend to look more like play-based social facilitation (peer-mediated interventions, joint attention training) than structured social skills curricula.

Rao and colleagues’ 2008 review in the Journal of the American Academy of Child and Adolescent Psychiatry found that social skills interventions with adolescents with higher cognitive ability and language levels showed the strongest and most durable gains. This does not mean lower-ability or non-speaking children cannot benefit — but the specific format and outcome targets need to match the individual.

Evaluating a Program’s Approach to Neurodiversity

Practical questions parents can ask before enrolling a child in social skills training:

Does the program have a manualixed protocol with published outcome data? Programs that cannot point to their evidence base deserve scrutiny.

Does the program involve parents systematically? Programs that deliver group sessions to youth without a parallel parent component are delivering a partial version of what the research studied.

How does the program address stimming and other autistic self-regulatory behaviors? A program that punishes or actively discourages stimming is teaching masking; one that contextualizes it (some stimming is fine privately and may be modulated in specific public contexts) is respecting autistic experience.

Does the program use peer generalization practice — real peers, not just role-play with adults? Does it include naturalistic settings as part of the program?

What outcomes does the program measure and report? Programs that only measure parent satisfaction are not measuring what the research measures.

For children whose social difficulties connect to emotional regulation, addressing both simultaneously produces better outcomes than treating social skills in isolation — dysregulation during social situations is one of the primary barriers to applying social knowledge in real time.

What to Watch for Over the Next 3 Months

If your child is enrolled in or beginning a social skills training program, track two categories: in-program performance and real-world transfer.

In-program, watch for whether your child can demonstrate the target skills in role-play by weeks 4–6. Most well-designed programs show rapid acquisition of knowledge; this is the easy part.

In real-world settings, watch for any spontaneous application of skills in naturalistic contexts — initiating with a familiar peer, navigating a turn-taking disagreement, joining an activity — without prompting. Early signs of generalization are positive indicators that the program is working the way the research says it should.

If generalization is absent at 8–10 weeks despite in-session gains, discuss with the program whether the homework and parent involvement components are being fully implemented. The generalization failure is usually a program delivery problem, not a child failure.

Watch also for signs of masking-related stress: increased exhaustion after social situations, more meltdowns at home after school, increased anxiety specifically around social events. These may indicate the program is producing effortful performance rather than genuine skill building.

Frequently Asked Questions

How long does social skills training need to last to produce results?

The PEERS program is 16 weeks, which is the best-studied duration. Shorter programs (8–10 weeks) may produce gains in social knowledge but have less evidence for behavioral generalization. Programs substantially shorter than 10 sessions have minimal evidence. Follow-up booster sessions at 6–12 months support maintenance.

Is group format better than individual social skills instruction?

For social skills specifically, group formats have an inherent advantage: the presence of peers creates naturalistic practice opportunities. Individual sessions can build knowledge and prepare a child for group participation but cannot replicate the complexity of actual peer interaction. Most evidence-based programs use group format with 4–8 participants.

My child has very high support needs. Is PEERS appropriate?

PEERS in its standard format requires sufficient cognitive and language ability to engage with the curriculum and apply explicit social rules. Children with higher support needs may benefit from adapted versions or from alternative approaches — peer-mediated interventions, naturalistic developmental behavioral interventions, or joint attention training — that do not require the same metacognitive demands. Consult with a behavioral psychologist about the appropriate fit.

Can social skills training cause harm?

Poorly delivered programs — particularly those that focus heavily on eliminating autistic behaviors without respecting the child’s wellbeing — can contribute to masking and the associated mental health costs documented in adult autistic populations. Well-delivered programs by trained clinicians who understand neurodiversity considerations are generally not harmful and are associated with meaningful gains. Program quality and clinician training matter enormously.

What can parents do at home to support generalization?

The PEERS parent group is explicitly designed to address this. Core components: helping your child identify 1–2 interested peers and create low-stakes interaction opportunities (organized around shared interests), practicing conversation skills in brief daily conversations at home, debriefing social situations supportively without judgment, and providing specific positive feedback when naturalistic skills are applied successfully.

How do I explain social skills training to my child without shaming them?

Frame the program around expanding options rather than fixing deficits. “This is a group where you’ll learn some strategies for connecting with kids who might be interested in the same things you are” positions it as skill expansion. Avoiding deficit framing — “you need to learn how to make friends,” “you’re not doing it right” — matters for engagement and for the child’s sense of self.

---

About the author

Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.

Sources

1. Laugeson, E. A., Frankel, F., Gantman, A., Dillon, A. R., & Mogil, C. (2012). “Evidence-based social skills training for adolescents with autism spectrum disorders: The UCLA PEERS program.” Journal of Autism and Developmental Disorders, 42(6), 1025–1036.
2. Laugeson, E. A., Gantman, A., Kapp, S. K., Orenski, K., & Ellingsen, R. (2015). “A randomized controlled trial to improve social skills in young adults with autism spectrum disorder: The UCLA PEERS® Program.” Journal of Autism and Developmental Disorders, 45(12), 3978–3989.
3. Schohl, K. A., Van Hecke, A. V., Carson, A. M., Dolan, B., Karst, J., & Stevens, S. (2014). “A replication and extension of the PEERS intervention: Examining effects on social skills and social anxiety in adolescents with autism spectrum disorders.” Journal of Autism and Developmental Disorders, 44(3), 532–545.
4. White, S. W., Keonig, K., & Scahill, L. (2007). “Social skills development in children with autism spectrum disorders: A review of the intervention research.” Journal of Autism and Developmental Disorders, 37(10), 1858–1868.
5. Cassidy, S., Bradley, L., Shaw, R., & Baron-Cohen, S. (2018). “Risk markers for suicidality in autistic adults.” Molecular Autism, 9(1), 42.
6. Lai, M. C., Lombardo, M. V., Ruigrok, A. N., Chakrabarti, B., Auyeung, B., Szatmari, P., … & Baron-Cohen, S. (2020). “Quantifying and exploring camouflaging in men and women with autism.” Autism, 21(6), 690–702.
7. Rao, P. A., Beidel, D. C., & Murray, M. J. (2008). “Social skills interventions for children with Asperger’s syndrome or high-functioning autism: A review and recommendations.” Journal of Autism and Developmental Disorders, 38(2), 353–361.
8. Gresham, F. M., Sugai, G., & Horner, R. H. (2001). “Interpreting outcomes of social skills training for students with high-incidence disabilities.” Exceptional Children, 67(3), 331–344.