This paper explores the relationship between adolescent marijuana use and school attendance. Data were pooled from the 1997 and 1998 National Household Surveys on Drug Abuse to form a sample of 15 168 adolescents, aged 12-18 years, who had not yet complete high school. The analysis determined the role of marijuana use in adolescent school dropout and, conditional on being enrolled, estimated the number of days truant. The potential endogeneity of marijuana use was tested in all specifications. The results indicate that any marijuana use was positively associated with school dropout and truancy in all models. However, when chronic marijuana use (weekly or more frequent) was distinguished from non-chronic marijuana use (less frequent than weekly), chronic marijuana use was found to be the dominant factor in these relationships. The results have important implications for educators, substance abuse treatment providers, and policymakers.