This study examined the effects of smoking policy on 4,807 adolescents in 23 schools over a two-county area in California. Amounts and prevalence rates of adolescent smoking were measured with a self-report survey and a biochemical measure; school smoking policy was measured with two independent surveys of school staff. Policy effects were evaluated with multiple and logistic regression analyses controlling for school-level socioeconomic status and environmental support for teaching and administration. Of the 23 schools, 100 percent had a formal written and regularly enforced policy component restricting student smoking on school grounds, 94 percent restricted students leaving school grounds, 65 percent restricted smoking near school grounds, and 57 percent had a smoking prevention education plan. Schools with policies having all four versus less than four components, high versus low emphasis on prevention, and a low versus high emphasis on cessation reported lower amounts of smoking in the last week and in the last 24 hours. Punitive consequences of policy violation had no effect. Results were compared to school staff observations of adolescent smoking, and school archival records of student smoking violations in the last year. Results suggest that school smoking policy is associated with decreased amounts of smoking in adolescents.