Background: Group-based psychotherapy in Autism Spectrum Disorder (ASD) has predominantly been studied in the United States by small studies in school-aged children without long-term follow-up. We report results of a large, confirmatory, multicentre randomized-controlled phase-III trial in children and adolescents studying the ASD specific, manualized group-based cognitive behavioural SOSTA-FRA approach. Methods: High-functioning ASD individuals aged 8-19 years old were randomized to 12 sessions SOSTA-FRA or treatment as usual. Primary outcomes were change in total raw score of the parent-rated Social Responsiveness Scale (pSRS) between baseline (T2) and end of intervention (T4), and between T2 and 3 months after end of intervention (T5). Trial registration: ISRCTN94863788. Results: Between 20/5/2010 and 14/2/2013, n = 320 ASD patients were screened, n = 228 patients were randomized, and N = 209 analysed. Mean pSRS difference between groups at T4 was -6.5 (95% CI -11.6 to -1.4; p = .013), and at T5 -6.4 (-11.5 to -1.3, p = .015). Pre-treatment SRS and IQ were positively associated with stronger improvement at T4 and T5. Conclusions: Short-term ASD-specific add-on group-based psychotherapy has shown postintervention efficacy with regard to parent-rated social responsiveness predominantly in male high-functioning children and adolescents with ASD. Future studies should implement blinded standardized observational measures of peer-related social interaction.