© 2014 Society for the Study of Addiction. BACKGROUND AND AIMS: Prenatal exposure to maternal cigarette smoking (PEMCS) is associated with a higher probability of substance use in adolescence. We explore if externalizing behavior mediates this relationship, while controlling for a number of potential covariates of this mediation process. METHODS: We used data obtained in two geographically distinct community samples of adolescents. The first (cross-sectional) sample consisted of 996 adolescents (12-18 years of age) recruited from the Saguenay Youth Study (SYS) in Canada (47% with PEMCS). The second (longitudinal) sample consisted of 1141 adolescents (49% with PEMCS) from the Northern Finland Birth Cohort (NFBC1986). In both samples, externalizing behavior and substance use were assessed during adolescence. In the NFBC1986 cohort, externalizing behavior was also assessed in childhood. RESULTS: In both populations, PEMCS is associated with a higher likelihood of adolescent drug experimentation. In the NFBC1986 cohort, exposed (versus non-exposed) adolescents experiment with an extra 1.27 [B = 0.24, 95% confidence intervals (CI) = 0.15, 0.33 P < 0.001] drugs. In the SYS cohort, a clear protective effect of not being exposed is shown: non-exposed (versus exposed) adolescents are 1.5 times [B = -0.42, 95% CI = -0.75, -0.09, P = 0.013] less likely to take drugs. These associations between PEMCS and drug experimentation remain in the multivariate and mediational analyses. CONCLUSIONS: Prenatal exposure to maternal cigarette smoking appears to be associated with a higher probability of experimenting with drugs during adolescence, both directly and indirectly via externalizing behavior and the number of peers reported as using drugs.