We tested whether marijuana use in the 2 months before the last menstrual period and during pregnancy affects the risk of spontaneous abortions of known karyotype. Spontaneous abortions (cases) were defined as chromosomally normal (n = 567) or chromosomally aberrant (n = 393) and, within the latter, by type of aberration (212 trisomies, 71 monosomies X, 49 triploidies, 61 others). Controls were women with prenatal care before 22 weeks gestation and delivering at 28 weeks or later (n = 2042). In comparison with controls, adjusted odds (OR) of reported marijuana use in chromosomally normal cases were 1.1 (95% confidence interval (CI) 0.7, 1.5) and in chromosomally aberrant cases combined 1.2 (95% CI 0.7, 1.9). With respect to specific aberrations, use in the perifertilisation period did not differ significantly from that in controls for trisomies (adjusted OR = 0.8, 95% CI 0.4, 1.8), monosomies X (adjusted OR = 1.8, 95% CI 0.7, 4.3), and triploidies (adjusted OR = 1.3, 95% CI 0.4, 4.5). Comparison of karyotype groups with each other yielded similar results. Our data do not support causal associations of marijuana use, at the levels represented in our sample, with either chromosomally normal or trisomic spontaneous abortion. With monosomy X and triploidy, no statistically significant associations were detected although numbers were insufficient to rule out moderate effects.