We conducted meta-analyses of relationships between highly prevalent substance use during pregnancy and offspring conduct disorder problems. In total 36 studies were included. Odds ratios (ORs) were 2.06 (1.67–2.54, 25 studies) for maternal smoking, 2.11 (1.42–3.15, 9 studies) for alcohol use, and 1.29 (0.93–1.81, 3 studies) for cannabis use, while a single study of caffeine use reported no effects. Our meta-analyses support an association between smoking and alcohol use during pregnancy, and offspring conduct problems, yet do not resolve causality issues given potential confounding by genetic factors, gene-environment interactions, and comorbidity such as with attention deficit hyperactivity disorders. Future studies should use genetically sensitive designs to investigate the role of pregnancy substance use in offspring conduct problems and may consider more broadly defined behavioral problems.