We evaluated the impact of nine community-based drug prevention, education, and treatment projects for pregnant and postpartum women using a quasi-experimental design. Projects provided case management and referral to services or provided day treatment. Self-reported measures of six substances were collected (a) from intake to delivery and (b) delivery to 6 months postpartum. We hypothesized that women who received project services (n = 370) between intake and follow-ups would be more likely to deter or reduce substance use than women who received an alternative or no intervention (n = 288). Data showed that project clients had significantly lower 30-day use rates on four of the measures-alcohol, any illicit drug(s), marijuana, and crack-from intake to delivery, with preintervention alcohol and other drugs use controlled. However, none of these results was maintained from intake through 6 months postpartum. Findings are discussed in terms of the difficulty of detecting and documenting promising intervention program effects over time in this population. (C) 2000 Elsevier Science Ltd.