Sign Up
Login
Try for Free
Login
The SRS-2 (Social Responsiveness Scale, Second Edition) is a 65-item rating scale quantifying autism-related social impairment from age 2 and a half through adulthood, completed by parents, teachers, and adults about themselves. Psychologists use it as a corroborating measure in autism evaluations and a progress measure in ABA programs. This page covers how to write up SRS-2 results, with a fictional sample and a results-section template.
Psychologists, school psychologists, and developmental pediatricians; publisher qualification level C
Autism evaluation teams, IEP teams, referring physicians, ABA programs and payers, parents and caregivers
150 to 400 words for the SRS-2 section · administration 15 to 20 minutes per form
Multi-informant autism trait rating scale
Autism diagnostic evaluations, ABA outcome measurement, school evaluations, adult and late-identification assessments, treatment monitoring
Published by WPS (2012); SRS-3 released in 2026; described here for write-up purposes, no test content reproduced
The Social Responsiveness Scale, Second Edition (SRS-2) is a 65-item, norm-referenced rating scale by John N. Constantino, MD, with Christian P. Gruber, published by WPS in 2012, that identifies autism-related social impairment and quantifies its severity across the lifespan. Four forms cover the age span: Preschool (ages 2.5 to 4.5) and School-Age (ages 4 to 18), each completed by a parent or teacher, plus an Adult form completed by a relative or other informant and an Adult Self-Report, both for ages 19 and up. Every form yields T scores (mean 50, standard deviation 10) for a Total score, two DSM-5 compatible subscales (Social Communication and Interaction; Restricted Interests and Repetitive Behavior), and five treatment subscales, against norms from 1,906 individuals and 4,709 ratings, separated by rater type and by the age and gender of the person rated. Scoring runs by hand on AutoScore forms or online through WPS's platform. Edition status matters as of July 2026: the SRS-3 published in 2026 with re-normed forms and structural changes, Pearson's Australian distributor page has carried an "SRS-3 Coming Soon in 2026" banner, and the SRS-2 remains on sale and in wide payer and program use during the transition.
The load-bearing distinction for write-ups: the SRS-2 measures a trait dimension, it does not decide a diagnosis. Its scores quantify how much autism-related social difficulty an informant observes, and elevations are consistent with, but not specific to, autism: peer-reviewed comparisons show strong separation of autism from typical development but substantially weaker separation from disruptive behavior disorders, social anxiety and selective mutism, and other conditions, and unadjusted scores can read as "general levels of impairment" rather than autism-specific severity. That is why the SRS-2 enters a psychological evaluation report or a diagnostic evaluation as one converging data stream alongside direct observation such as the ADOS-2, developmental history, and clinical judgment. A results section that reports the scores dimensionally, addresses what else could elevate them, and lets the whole evaluation carry the diagnostic conclusion is most of what separates a defensible SRS-2 write-up from a score printout with sentences around it.
Psychologists and school psychologists write up SRS-2 results inside autism diagnostic evaluations and school evaluations, where it supplies the quantified, multi-informant view of everyday social behavior that direct observation cannot; developmental pediatricians and psychiatrists order it for diagnostic clarification inside a diagnostic evaluation; ABA programs administer it on a schedule because TRICARE's Autism Care Demonstration requires the Parent Form at baseline and every year as an outcome measure; and clinicians assessing adults use the Adult and Adult Self-Report forms in late-identification workups, where the self-report has published validation in autistic adults. It is the right tool when the question is dimensional: how much autism-related social difficulty do the people around this person observe, in which settings, and how has it changed. The instruments around it then carry their own loads: the ADOS-2 for structured direct observation, the ADI-R for the developmental interview, a broadband scale such as the BASC-3 when the referral question spans attention, mood, and conduct rather than social reciprocity alone, and adaptive measures such as the Vineland-3 for daily-living skills. The SRS-2 write-up is where the informant evidence gets integrated: multiple raters, one dimensional account, clearly subordinated to the full evaluation.
No statute, payer, or publisher mandates a results-section format. The sequence below is the convention experienced evaluators converge on because it survives review: it names what was administered and who the raters were, reports the Total score before the subscales, integrates informants instead of averaging them, and closes with the specificity statement that this instrument, more than most, cannot do without. Each section carries the pitfall that most often undermines it.
Measures, forms, raters, and norms. Name the instrument and edition, the form completed by each rater (Preschool, School-Age, Adult, Adult Self-Report), each rater's relationship and setting, completion dates, and the norm reference, since SRS-2 norms are separated by rater type and by the age and gender of the person rated. Pitfall: writing "an SRS" with no edition or form. With the SRS-3 published in 2026 and re-normed, an unnamed edition leaves reviewers unable to tell which norms produced your T scores.
Rating quality and completeness. The SRS-2 has no formal validity indexes, so the write-up carries that weight in prose: state that forms were complete, each rater's opportunity to observe, and anything that could color the ratings, such as rater stress or a recent crisis. Pitfall: treating informant ratings as objective measurement. They are structured observations from a vantage point, and the new Inconsistent Responding Index in the SRS-3 exists precisely because rating quality varies.
Total score first. Report the Total T score with its qualitative severity descriptor in words (within normal limits, mild, moderate, severe) and its plain-language meaning. The Total score is the instrument's primary, continuously distributed index of severity. Pitfall: reporting a numeric threshold as if it were a diagnostic cutoff. The descriptor bands characterize normative elevation, and no T score establishes or excludes a diagnosis.
DSM-5 compatible subscales. Report Social Communication and Interaction and Restricted Interests and Repetitive Behavior, which map onto the two DSM-5 criterion domains and have factor-analytic support in a sample of more than 9,000. Pitfall: building the interpretation on the five treatment subscales instead. The publisher states they are not used for screening or diagnosis; they are intervention-planning aids, and reviewers know the difference.
Cross-informant integration. Report each informant's scores and interpret convergence and divergence rather than summarizing forms in sequence. Parent and teacher agreement is modest by design across instruments, averaging r = .28 in a meta-analysis of 341 studies, and the documented SRS-2 conventions are to treat an elevation from either informant as a flag for full evaluation and to seek additional rater input when scores differ by more than 10 T points (Texas Statewide Leadership for Autism Training). Pitfall: averaging raters into a single number, or crowning one rater correct. Divergence is usually setting information, and the write-up should say what it means.
Differential and specificity statement. One sentence stating that elevations are consistent with, but not specific to, autism, followed by the contributors you considered: attention and behavior problems, anxiety, language and cognitive level, trauma. Published comparisons show weaker discrimination from disruptive behavior disorders and overlap with social anxiety and selective mutism. Pitfall: omitting the statement entirely. This is the single most common substantive gap reviewers flag in SRS-2 sections, because the instrument's documented weakness is specificity, not sensitivity.
Integration with direct measures and history. State where the SRS-2 findings converge with or diverge from direct observation (such as the ADOS-2), the developmental interview, and records. Correlations between the SRS-2 and direct diagnostic measures are moderate at best, so some disagreement is expected. Pitfall: explaining divergence away. A rating-scale elevation the observation did not corroborate, or the reverse, is a finding to interpret, not an inconvenience to bury.
Interpretive summary and recommendations linkage. Answer the referral question dimensionally, state what the scores do and do not establish, and tie each recommendation to a finding, including the re-rating plan: who completes which form, when, and why, matching any program cadence such as TRICARE's annual requirement. Pitfall: recommendations that could follow any profile. If nothing in the plan traces to the informant evidence, either the plan or the results section is incomplete.
MEASURES, FORMS, AND RATERS Instrument/edition: SRS-2 Form(s): [Preschool / School-Age / Adult / Adult Self-Report] Raters: [name or role, relationship, setting, date completed] Norm reference: [rater type; age and gender of person rated] RATING QUALITY [Forms complete? Each rater's opportunity to observe; factors that could color ratings; how any concern was handled] TOTAL SCORE [Total T per rater + qualitative descriptor in words + plain-language meaning; no numeric cutoffs] DSM-5 COMPATIBLE SUBSCALES Social Communication and Interaction: [T scores, descriptor] Restricted Interests and Repetitive Behavior: [T scores] Treatment subscales (if reported): [intervention planning only] CROSS-INFORMANT INTEGRATION [Where raters converge; what divergence means (setting, perspective); >10 T-point difference: note added input] DIFFERENTIAL AND SPECIFICITY STATEMENT [Elevations consistent with, not specific to, autism; contributors considered: ADHD, anxiety, language, other] INTEGRATION WITH DIRECT MEASURES AND HISTORY [Convergence/divergence with observation, interview, records] INTERPRETIVE SUMMARY (answer the referral question) [Dimensional statement; what the scores do NOT establish] RECOMMENDATIONS LINKAGE [Each recommendation tied to a finding; re-rating plan: form, rater, timing (payer cadence if applicable)] Evaluator signature / credentials: Date:
Free to use and share, no signup. The PDF includes a one-page cheat sheet with section-by-section pitfalls and a pre-sign checklist; the DOCX is the blank results-section skeleton, ready to adapt.
Scenario: an 8-year-old referred by his pediatrician for an autism diagnostic evaluation after two years of social difficulties at school. Parent and teacher School-Age forms diverge by more than 10 T points, the cross-informant shape where the write-up's job is to interpret the difference instead of averaging it. This is the SRS-2 section of the larger evaluation report only, condensed but structurally complete. All details are fictional.
Client: L.T., 8 · Referral: autism diagnostic evaluation, referred by pediatrician · Evaluator: M. Okafor, PsyD, Licensed Psychologist · Ratings completed: 07/06/2026 (parent), 07/08/2026 (teacher) · Report date: 07/16/2026
Measures and administration: The Social Responsiveness Scale, Second Edition (SRS-2) School-Age Form was completed independently by L.T.'s mother and by his third-grade classroom teacher as part of an evaluation that also included the ADOS-2, a developmental history interview, cognitive screening, and record review, reported in their own sections. Scores are T scores (mean 50, SD 10) against the SRS-2 norms for each rater type and for L.T.'s age and gender. Severity descriptors below follow the manual's qualitative conventions and are stated in words; no score is treated as a diagnostic threshold.
Rating quality: Both forms were fully completed. L.T.'s mother sees him daily across home and community settings; his teacher has observed him across seven months of structured classroom, lunchroom, and recess contexts. No recent family disruption or rater concern was identified that would be expected to color the ratings, and both raters described their responses as reflecting typical recent behavior.
| SRS-2 scale | Parent T score | Teacher T score |
|---|---|---|
| Total | 78 | 66 |
| Social Communication and Interaction (DSM-5) | 77 | 66 |
| Restricted Interests and Repetitive Behavior (DSM-5) | 74 | 63 |
| Social Awareness (treatment) | 68 | 58 |
| Social Cognition (treatment) | 74 | 64 |
| Social Communication (treatment) | 76 | 65 |
| Social Motivation (treatment) | 70 | 61 |
Total score: L.T.'s mother's ratings produced a Total T score of 78, in the range the manual describes as severe, indicating that she observes social difficulties of a kind and frequency strongly associated with clinically significant interference in everyday interaction. His teacher's ratings produced a Total T score of 66, in the moderate range, indicating clearly elevated but less pervasive difficulty in the classroom setting.
DSM-5 compatible subscales: Both raters elevated both domains. Social Communication and Interaction (parent T = 77; teacher T = 66) and Restricted Interests and Repetitive Behavior (parent T = 74; teacher T = 63) were consistent with the Total score pattern for each rater, indicating that the reported difficulty spans both DSM-5 criterion domains rather than concentrating in one. Treatment subscale scores followed the same gradient for both raters, with relative strength in Social Awareness, which his teacher rated within normal limits; these subscale scores are used here for intervention planning only.
Cross-informant integration: The 12-point Total score difference exceeds the 10-point convention for seeking additional perspective, and additional input was obtained from L.T.'s after-school program coordinator by interview, which described social behavior closer to the mother's account in unstructured settings. The divergence is interpreted as setting information: the classroom's predictable structure and adult scaffolding appear to reduce the visible load of L.T.'s social difficulty, while less structured home and community settings expose it. Both raters' elevations independently exceed the level treated as a flag for full evaluation, and neither is treated as more correct.
Differential and specificity statement: Elevated SRS-2 scores are consistent with, but not specific to, autism; attention problems, anxiety, language weaknesses, and disruptive behavior can each raise scores. Rating-scale and interview data in this evaluation showed no attentional or oppositional pattern of that kind, language screening was age-typical, and the anxiety symptoms described were situational and secondary to social demands, so these contributors do not account for the elevations observed.
Integration with direct measures and history: The SRS-2 pattern converges with the ADOS-2 observation, reported separately, and with a developmental history of social-communication differences evident before age 3. Cross-informant rating-scale evidence, direct observation, and history therefore point in the same direction, and the SRS-2 findings strengthen, but do not by themselves establish, the diagnostic formulation presented in the summary of this report.
Recommendations linkage: The cross-setting social-communication findings support the social-communication intervention goals and the structured-peer-interaction recommendation in the plan; the relative strength in awareness supports building on motivation-side supports rather than insight-first approaches. Re-rating with the SRS-2 School-Age Form by the same two raters in 12 months is recommended to monitor change on the same instrument, edition, and norms, aligned with the program's annual outcome-measure cadence.
This sample is fictional and for educational purposes. It does not describe a real client or record, and the scores are invented for illustration and correspond to no real child or record.
Writing these after every session? BastionGPT drafts complete notes from bullets, dictation, or a transcript.
Generate a note from bulletsWrite the SRS-2 section knowing which program will read it, because the instrument's status changes sharply by payer and jurisdiction. Under TRICARE's Autism Care Demonstration it is PAYER POLICY at its most explicit: the TRICARE Operations Manual requires the SRS-2 Parent Form at baseline and every year thereafter as an outcome measure alongside the Vineland-3 and the PDDBI, treats ABA-provider-completed measures as indirect services under 97151, and sets a documentation rule worth taping to the wall: submissions must include the full publisher print report or hand-scored protocol with summary score sheets, because "Imbedding T-scores within the treatment plan or other clinical documents is insufficient." Georgia Medicaid's pathway, as implemented in Peach State's clinical policy GA.CP.BH.504 (revised October 2025), requires a minimum of two assessment tools, one clinician-administered and one caregiver tool, and lists the SRS-2 among the accepted caregiver tools; the evaluation report must summarize each instrument with scores, dates, and the evaluator's signature and credentials. Most commercial policies run the other way: Aetna's clinical policy CPB 0648 names the ADI-R, ADOS-2, CARS-2, and Asperger Syndrome Diagnostic Scale as diagnostic tools and does not name the SRS-2, so an SRS-2-forward report can read as thin to a reviewer working from that list. In US schools, the LAW frame is IDEA's instrument-neutral rule that evaluators may "not use any single measure or assessment as the sole criterion" (34 CFR 300.304(b)(2)). Testing and rating-scale work is billed inside the psychological and developmental testing code families by bare code number (for example 96110, 96112, 96127 for screening instruments; 96130, 96131, 96136 through 96139 for testing services), which is payer policy and coding convention, not law. In Canada, British Columbia's Autism Funding pathway states that for a private diagnosis "both the ADOS-2 and ADI-R are required instruments" (MCFD form CF0904), so the SRS-2 supports but never qualifies a BC funding application, and the Canadian Paediatric Society's diagnostic standards treat rating scales as adjuncts, noting that findings from a diagnostic tool "cannot be used alone to diagnose ASD" while observing that, uniquely on its list, training is not required to use the SRS-2. In Australia, the National Guideline (2018, second edition 2023, NHMRC-approved recommendations) describes a comprehensive assessment process without mandating any specific instrument, the NDIS asks for diagnosis plus functional-capacity evidence, and the MBS names no instruments at all (a July 2026 search of MBS Online for the scale returns no items), while record rules come from the Psychology Board's code of conduct: reports true and objective, records retained seven years, or until age 25 for child clients.
Edition currency became a live question in 2026. The SRS-3 published in 2026 with a new normative sample of 1,180 individuals and 2,340 ratings, a new School-Age Self-Report Form, three norm options for every form (female, male, or all-sample), the Restricted Interests and Repetitive Behavior treatment subscale split into Cognitive Flexibility and Sensory-Motor Behaviors for six treatment subscales in all, a new Inconsistent Responding Index to flag random response patterns, and Spanish forms available online only. During the transition the defensible practices are specific: name the edition and norm reference in every report; finish an evaluation on the edition it started on; keep a monitoring sequence on one edition and norms, since a re-rating on new norms is not comparable to a baseline on old ones and any cross-edition comparison belongs in the text as a stated limitation; and check payer language before switching, because while TRICARE's manual entry already reads "SRS-2 (or Current Edition)", contractor submission systems and state tool lists such as Georgia's still name the SRS-2, and independent peer review of the SRS-3 will lag its launch. The qualification boundary is the publisher's: both editions are qualification level C, and WPS states that it "only qualifies individuals for test purchase, not for test use", which leaves use governed by licensing boards and scope of practice; forms are completed by laypeople, but interpretation and write-up are professional acts. One more tension belongs in the writer's head rather than on the page: the publisher's marketing describes the SRS-2 as offering "the convenience of a screener and the power of a diagnostic tool", while the peer-reviewed record documents weak discrimination from other clinical conditions, and a defensible report writes to the evidence, reporting elevations dimensionally and letting the full evaluation carry the diagnosis.
SRS, SRS-2, and SRS-3 are trademarks of Western Psychological Services (WPS). BastionGPT is not affiliated with, or endorsed by, the publisher. This page reproduces no test items, stimuli, norms, or scoring materials.
There is no payer audit series for rating-scale write-ups; the accountability record here is psychometric, plus one payer paper trail that is unusually explicit. The peer-reviewed findings that matter: the SRS separates autism from typical development almost perfectly but discriminates far less well from disruptive behavior disorders, it overestimated autism in boys with selective mutism, the Adult Self-Report showed poor discriminant validity against anxiety, and unadjusted scores can index "general levels of impairment" rather than autism-specific severity. On the payer side, TRICARE's ACD submission rules reject write-ups whose scores exist only in the narrative. Those findings predict almost every failure below. The BastionGPT Clinical Advisory Board sees the same errors most often in SRS-2 write-up reviews:
BastionGPT is specifically trained, tuned, and clinically tested on psychological and psychoeducational evaluation reports.
See how clinicians use it day to day on the AI therapy notes page.
Many BastionGPT users report saving more than 90 minutes per day on documentation.
HIPAA-compliant with a signed BAA on every plan. Your data is never used to train models. BastionGPT drafts, you review and sign.
Every SRS-2 form yields T scores, standard scores with a mean of 50 and a standard deviation of 10, for a Total score, two DSM-5 compatible subscales, and five treatment subscales, referenced to norms separated by rater type and by the age and gender of the person rated. Higher scores mean the rater observes more autism-related social difficulty. The severity descriptors (within normal limits, mild, moderate, severe) are qualitative characterizations of normative elevation, subject to measurement error and rater perspective. Write them as descriptors in words, pair each with its plain-language meaning, and let any diagnosis come from the whole evaluation, never from a score.
Both are current as of July 2026. The SRS-3 published in 2026 with a new normative sample (1,180 individuals, 2,340 ratings), a new School-Age Self-Report Form, three norm options for every form (female, male, or all-sample), the Restricted Interests and Repetitive Behavior treatment subscale split into Cognitive Flexibility and Sensory-Motor Behaviors, a new Inconsistent Responding Index, and Spanish forms online only, while the SRS-2 remains on sale and named in payer documents. A report on either edition is defensible if it names the edition and norm reference, finishes on the edition it started on, and treats any cross-edition score comparison as a stated limitation. Before a program-wide switch, check payer language: TRICARE's manual entry reads "SRS-2 (or Current Edition)", but contractor systems and state tool lists still name the SRS-2, and independent peer review of the SRS-3 will lag its launch.
No law or diagnostic standard names it, and its status varies by program. Where it is required, it is payer policy: the TRICARE Autism Care Demonstration requires the SRS-2 Parent Form at baseline and every year as an outcome measure for ABA services. Where it is accepted, it is one option: Georgia Medicaid's pathway (GA.CP.BH.504) lists it among caregiver tools in a required two-tool minimum. Where it is unnamed, other instruments are: Aetna's policy names the ADI-R, ADOS-2, and CARS-2, and British Columbia's Autism Funding pathway requires the ADOS-2 and ADI-R for a private diagnosis. In schools, IDEA's rule is instrument-neutral: no single measure may be the sole criterion (34 CFR 300.304(b)(2)), and Australia's National Guideline mandates no instrument at all. The honest report sentence: the SRS-2 is never sufficient, sometimes required as an outcome measure, and often not named at all.
Neither, and the write-up should not choose. Agreement between different informant types is modest across instruments, averaging r = .28 in a meta-analysis of 341 studies, and parents often rate higher than teachers. The documented SRS-2 conventions, summarized by the Texas Statewide Leadership for Autism Training: treat an elevation from either informant as a flag for full evaluation, and obtain additional rater input when scores differ by more than 10 T points. Divergence usually carries setting information, structure and scaffolding in one environment, demands and exposure in another, and the defensible paragraph reports both scores, interprets the difference, and refuses to average it.
No, and this is the instrument's best-documented limitation. The SRS separates autism from typical development almost perfectly, but in matched clinical comparisons discrimination fell substantially against disruptive behavior disorders, it overestimated autism in boys with selective mutism, and the Adult Self-Report showed poor discriminant validity against anxiety. Influential work concluded that without accounting for behavior problems, age, and language, scores can index "general levels of impairment" rather than autism-specific severity. The Canadian Paediatric Society's standards say it directly: findings from a diagnostic assessment tool "cannot be used alone to diagnose ASD". Write elevations as consistent with, but not specific to, autism, name the differential contributors you considered, and let observation, interview, and history carry the conclusion.
By method and job. The SRS-2 is an informant and self-report rating scale: graded-frequency items, lifespan forms, and a severity-quantifying T score, which makes it the dimensional, cross-setting layer. The ADOS-2 is a clinician-administered, semi-structured direct observation, the diagnostic anchor for most pathways. The CARS-2 is a clinician rating completed from observation and information, the ASRS is a DSM-aligned multi-informant rating system, and the SCQ is a brief yes-or-no screener. The battery convention in autism evaluations pairs a dimensional rating scale (SRS-2) with direct observation (ADOS-2), a developmental interview (ADI-R), and cognitive, language, and adaptive measures, and no instrument on that list substitutes for the comprehensive evaluation or for clinical judgment.
Not on this page, and not legitimately on any public page. The 65 items, scoring keys, and norm tables are protected test materials: the SRS-2 is qualification level C, WPS states that it "only qualifies individuals for test purchase, not for test use", and psychologists carry an ethical duty to maintain test security under APA Ethics Standard 9.11. Consumer sites offering SRS-2 self-scoring undermine exactly the controls that make the norms meaningful. What a report can include: T scores, severity descriptors in words, and your own prose describing what the scales measure. What it cannot: item text, scoring rules, or reproduced norm tables.
Yes. The Adult Self-Report form covers ages 19 and up and has published validation in a sample of 237 autistic adults, which matters in late-identification evaluations where childhood informants are scarce. Interpret it with two cautions. First, self-reported elevations overlap heavily with anxiety: the Adult Self-Report showed poor discriminant validity between autistic and highly anxious adults (the figure usually quoted from that study is a sensitivity value from that clinical comparison, not a specificity in anxious adults). Second, current measures and procedures may miss some autistic adults, particularly women, whose observable presentation can understate their internal experience. Pair the self-report with the Adult relative or other-informant form, developmental history, and direct assessment, and say in the write-up how the streams converged.
Yes. BastionGPT is trained and clinically tested on psychological and psychoeducational evaluation reports, the parent documents SRS-2 sections live inside. Paste a score summary (raters, forms, T scores, descriptors) and it drafts the results-section narrative with the cross-informant integration and the specificity statement in place for your review; it can also check a finished section for score-descriptor mismatches, a missing differential sentence, or an unnamed edition, and produce a plain-language summary for parents and teachers. BastionGPT is HIPAA-compliant with a signed BAA on every plan, your data is never used to train models, and drafting from scores you paste means no protocol or item content ever needs to leave your records.
The instrument facts and compliance claims on this page trace to these sources, last verified July 2026:
Educational content, not legal or billing advice. Sample notes are fictional. Follow your organization's policies and your board, payer, and jurisdiction requirements.