This study examined the association of depressive symptoms of adolescents and their smoking behavior. A national cohort sample of school adolescents (N = 5,855) who participated in the 1989 and 1993 Teenage Attitudes and Practices Survey was obtained. The information from this survey included measures of smoking behavior and a series of factors related to depressive symptoms during a 3-yr. span. A cross-lagged analysis with Kendall tau b correlations was used. Judging from the magnitude of the cross-lagged correlations, the direction of causation cannot be ascertained. Rather, a reciprocal relationship between depression and smoking may be suggested.