We examined the longitudinal impact of self-reported exposure to pro- and anti-tobacco media on adolescents' susceptibility to smoking, using in-school surveys from a culturally diverse sample. Ethnicity and acculturation also were examined as potential moderators. Middle-school students (N = 2,292) completed self-report questionnaires during the 6th, 7th, and 8th grades. Chi-square analyses were conducted to determine whether reported exposure to pro- and anti-tobacco media varied according to ethnicity, acculturation, and immigration status. Logistic regression models were used to examine whether pro- and anti-tobacco media exposure in 6th grade was associated with susceptibility to smoking by later grades. Recall of people smoking in television programs and pro-tobacco advertisements in stores was associated with adolescent smoking susceptibility. Exposure to anti-tobacco advertisements on television protected against susceptibility. No significant interaction effects between pro- and anti-tobacco media exposure on smoking susceptibility were found. Ethnicity and acculturation did not moderate these associations. Our longitudinal study provides evidence that pro-tobacco media and advertising increases susceptibility to smoking over time. More important, anti-tobacco advertisements are not sufficient to reduce the harmful effects of adolescent exposure to pro-tobacco media. Policy-level interventions such as restrictions in tobacco advertising may be necessary to prevent adolescent smoking.