Thirty-four schools (n=7426 consented sixth graders, 71% of the eligible population) were randomized to conditions to test the hypothesis that Skills for Adolescence (SFA), a widely used comprehensive life skills training curriculum with a dedicated drug education unit, is more effective than standard care in deterring and delaying substance use through middle school. Two-year posttest (1-year post-intervention) data were collected from 5691 eighth graders (77% of those who completed the sixth-grade survey and 87% of those who completed the seventh-grade survey). Lifetime and recent (last 30 days) use of five substances or combinations of substances was compared using mixed-model regression to control for school clustering. There were two significant treatment main effects at the end of the eighth grade: lifetime (P=.05) and recent (P<.03) marijuana use were lower in SFA than control schools with pretest usage and salient demographic and psychosocial variables controlled. There was also one significant Treatment x Pretest Usage interaction around binge drinking. Baseline binge drinkers in SFA schools were less likely to report recent binge drinking than students in control schools (P<.01); there were no treatment differences among baseline nonbinge drinkers. Analyses of potential mediators of SFA treatment effects on eighth-grade binge drinking and marijuana use suggested that SFA increased self-efficacy around drug refusal skills, but did not affect behavioral intentions, perceptions of harm, or perceived peer norms. These 2-year (1-year post-intervention) outcomes offer some additional support for SFA effectiveness and the general thrust of school-based, life skills-based prevention programs. The promising sixth- through eighth-grade findings for SFA, a commercially available program, provide a further step in bridging a major gap in the "research to practice" literature: theory-based interventions that have documented behavioral effects have not enjoyed large-scale implementation, while intuition-based programs that have no documented effects still enjoy wide exposure.