This article includes two comments on the article by G B Bovasso regarding cannabis abuse as a risk factor for depressive symptoms. D H Epstien expresses concern about Bovasso’s use of odds ratios in a longitudinal cohort, because they may have led to an exaggeration of relative risk. W Kuo also expresses concerns about the method for determining the risk set for the incidence of suicidal ideation in the subject population. Bovasso responds to these concerns, & disagrees with the arguments they pose.