This study assessed the risk of smoking uptake over 2 years in adolescent boys (in grades 7 and 9) who had used smokeless tobacco (ST). We used logistic regression to determine whether the odds of adolescent boys taking up regular smoking over a period of 2 years were greater among initial nonsmokers who used ST, compared with nonusers of ST, after accounting for six well-established predictors of smoking. Initiation of weekly smoking 2 years after baseline was associated with ST use at baseline, even after including dichotomous measures of parent, sibling, or close friend smoking; low academic grades; 30-day alcohol use; and a scale measure of deviant behavior. With these other predictors included in the model, the odds ratio for the association of ST use with weekly smoking after 2 years was strong and significant (OR = 2.55, 95% CI 1.45-4.47, p<.001). The use of ST in the 7th and 9th grades is a significant risk factor for subsequent smoking even when controlling for other factors.