This paper discusses a study that examines the relationship between participation in religious youth groups, smoking behavior and subsequent use of cocaine, marijuana, and alcohol. The document briefly surveys earlier research on smoking and drug abuse. The study surveyed students, schools, parents, teachers, and student academic achievement data. Surveys began with eighth grade students and followed up the same students 2 years later. The document presents three models of drug use. The first relates smoking behavior and participation in a religious youth group to predict subsequent marijuana use. The hypothesis is that variables occur before smoking behavior and that smoking is a gateway drug leading to marijuana use. The second model predicts cocaine use from the same variables. The same hypothesis applies. The third model predicts binge drinking with the same hypothesis and variables. Path analysis was particularly relevant to this study because the technique has the ability to decompose effects into direct, indirect, and total effects. This enables the researcher to identify links between the variables and to establish weak causal ordering. The results of the study show that religious participation was related to lower substance abuse. The findings also confirm the notion that smoking usually precedes substance abuse. The results of this study reiterate the importance of drug prevention programs and the fact that drug prevention activities should occur in more places than the public school. Programs designed to discourage drug use must also contain a component that deals with tobacco use as well. (DK)