objective: to determine the prevalence of the heavy use of drugs among elementary and high school students in a sample of public and private schools, and to identify associated demographic, psychological, cultural and social factors. methods: this report describes a cross-sectional study using an intention-type sampling technique that compared public schools in central and peripheral areas and private schools. an anonymous self-administered questionnaire was applied. the sample consisted of 2,287 elementary and high school students in the city of campinas in 1998. heavy use of drugs was defined as the use of drugs on 20 or more days during the 30 days preceding the survey (who, 1981). for the statistical analysis, polytomic logistic regression analysis (logit model) was utilized to identify factors that influenced this manner of using drugs. results: heavy use of legal and illegal drugs was found as follows: alcohol (11.9%), tobacco (11.7%), marijuana (4.4%), solvents (1.8%), cocaine (1.4%), medications (1.1%) and ecstasy (0.7%). the heavy use of drugs was greatest among students at the city-center public school who had daytime jobs and studied in the evenings. these students were in the a and b socioeconomic classes and had had little religious education during childhood. conclusions: greater availability of cash and specific socialization patterns were identified as factors associated with the heavy use of drugs among students.