We conducted a prospective study of the effects of marijuana use in 35 pregnancies. Thirty-six age- and parity-matched nonuser pregnancies were selected for comparison. (Users of other illicit drugs were excluded from study.) Users and nonusers were similar with respect to most potentially confounding prenatal risk factors, although users tended to come from lower income and educational background. Most adverse outcomes of pregnancy were too infrequent to allow reliable comparisons between the groups, but the infants born to users exhibited significantly more meconium staining (57%, versus 25% in nonusers). Significant differences in duration of labor were also observed. Possible reasons for these findings are discussed.