The aim of this study was to measure the prevalence of marijuana use in the last month, its associated factors and its relationship with academic migration among undergraduate students of a federal university in southern Brazil. This was a cross-sectional study and data were collected through self-administered questionnaire. A systematic sampling process was conducted. To data analyses, it was used Poisson regression with robust adjust for variance. Overall, 1,423 students participated. The prevalence of marijuana use in the last month was 16.8% (95%CI 14.8% to 18.8%). Data showed that the greater the distance of the city prior to university entry, the higher the prevalence of marijuana use in the last month. Being male, having less age, being single, not having religious practices, having relatives and friends who have used any illicit drug, and having tobacco use in the last month were also risk factors. We understand that academic migration within the country tends to decrease students contact with family and increase vulnerability to peer influence, which may lead to a higher probability of marijuana use in this sample. The results highlights the importance to develop projects of illicit drug use prevention focused on this risky subgroup.