The original study sample included 1,634 adolescents in the 7th, 8th and 9th grades in 11 Los Angeles County, Calif., schools. Data were obtained from 192 males and 462 females who participated in years 5 and 9 of a longitudinal study of adolescent development and drug use. It was hypothesized that social conformity in high school would predict both social conformity and less substance use in young adulthood; substance use in high school would predict substance use and lack of social conformity in young adulthood; and males and females would have similar structural associations between the latent variables. Men and women were analyzed separately and then contrasted directly. Findings indicate that social conformity decreased hard drug use for both males and females and decreased alcohol use for females. Social conformity did not affect cannabis use over time. Early substance use did not predict less later social conformity. Magnitude of structural paths between latent variables were reasonably similar for both males and females for hard drug use, more similar for alcohol use, and quite similar for cannabis use. The social implications of the differential acceptability of substance use by sex are discussed. Figures, tables, and 22 references. (Author abstract modified)