In 1990 the Kaiser Permanente Medical Care Program in northern California instituted a plan, based on obstetric clinics, to intervene actively in perinatal substance abuse. In addition to screening and early identification of any abuse problems, the program emphasizes early intervention, ongoing counseling, and case management by a clinical therapist located in the prenatal clinic, the Early Start Specialist, who is experienced in substance abuse. A retrospective cohort study was undertaken to assess this effort in 6774 mothers and their infants delivered from mid-1995 to mid-1998. Women found to be at increased risk of using alcohol, tobacco, or other agents during pregnancy were referred to the specialist who carried out an in-depth psychosocial assessment and formulated a follow-up management plan that included counseling when judged to be indicated. Study groups included 782 women who were screened, assessed, and treated (SAT), 348 who were screened and assessed (SA), 262 who were only screened (S), and 5382 control women without apparent substance abuse. Women in all the substance abuse groups began perinatal care after 13 weeks' gestation more often than did control women, but rates were significantly lower for the SAT than for the SA group. The total amount of prenatal care was highest for the SAT group. The SAT women were likelier than those in the SA group to have a positive toxicology screen while pregnant. Unadjusted rates of assisted ventilation, low birth weight, and preterm delivery were lowest in the SAT and control groups. The SA and S groups had significantly higher rates than control women for these outcomes, and the SAT group had lower rates than the SA and S groups. None of 24 neonatal deaths occurred in the SAT group, but there were no significant group differences in neonatal mortality rates. Multivariate logistic regression analyses adjusting for numerous factors (maternal age, ethnicity, number of prenatal visits, coming late to prenatal care) yielded similar results. Odds ratios for assisted ventilation, low birth weight, and preterm delivery were elevated in the SA and S groups compared with control women. Admission to neonatal intensive care was significantly more frequent for all the substance abuse groups. This experience points to the need to improve perinatal outcomes in nonacademic settings, and also the value of cooperative efforts by clinicians and mental health professionals. It seems wise to closely integrate substance abuse treatment with regular prenatal care.