Aims. We sought to estimate the contribution of genetic and environmental factors to adolescent tobacco, alcohol and other substance use. Design, setting and participants. The sample consisted of 327 monozygotic and 174 like-sex dizygotic twin pairs born in Minnesota and aged 17-18 years at time of assessment. Biometrical methods were used to estimate the contribution of additive genetic, shared and non-shared environmental factors to adolescent substance use. Measurements. As part of a day-long psychological assessment, adolescent twins completed a computerized substance use interview to determine whether they had ever used tobacco, alcohol or other illicit drugs. Findings. The heritability for the liabilities to tobacco, alcohol and other drug use was estimated to be 59%, 60% and 33% among males, and 11%, 10% and 11% among females. However, the gender difference was not statistically significant. Estimates of shared environmental effect were substantial and insignificantly higher among females (71%, 68% and 36%, respectively) than among males (18%, 23% and 23%, respectively). The covariation among the three substance use phenotypes could be accounted for by a common underlying substance use factor. Estimates of the contributions of genetic, shared environmental and non-shared environmental factors to variance in this factor were 23% 63% and 14%, respectively. Conclusions. These findings add to the growing behavioral genetic literature indicating that adolescent initiation of substance use, a powerful predictor of adult substance use diagnosis, is influenced primarily by environmental rather than genetic factors.