OBJECTIVE: The aim of this study was to test whether a cognitive-behavioral and mindfulness-based group sleep intervention would improve behavior problems in at-risk adolescents, and whether these improvements were specifically related to improvements in sleep. METHOD: Secondary analysis of a randomized controlled trial conducted with 123 adolescent participants (female = 60%; mean age = 14.48, range 12.04-16.31 years) who had high levels of sleep problems and anxiety symptoms. Participants were randomized into either a sleep improvement intervention (n = 63) or an active control "study skills" intervention (n = 60). Participants completed sleep and behavior problems questionnaires, wore an actiwatch and completed a sleep diary for five school nights, both before and after the intervention. RESULTS: Parallel multiple mediation models showed that postintervention improvements in social problems, attention problems, and aggressive behaviors were specifically mediated by moderate improvements in self-reported sleep quality on school nights, but were not mediated by moderate improvements in actigraphy-assessed sleep onset latency or sleep diary-measured sleep efficiency on school nights. CONCLUSION: This study provides evidence, using a methodologically rigorous design, that a cognitive-behavioral and mindfulness-based group sleep intervention improved behavior problems in at-risk adolescent by improving perceived sleep quality on school nights. These findings suggest that sleep interventions could be directed towards adolescents with behavior problems. CLINICAL TRIAL REGISTRATION: This study was part of The SENSE Study (Sleep and Education: learning New Skills Early). URL: ACTRN12612001177842; http://www.anzctr.org.au/TrialSearch.aspx?searchTxt=ACTRN12612001177842&isBasic=True.