Home visitation interventions show promise for helping at-risk mothers, yet few programs have been developed and evaluated specifically for alcohol and drug-abusing pregnant women. This study examines outcomes among 216 women enrolled in the Washington State Parent-Child Assistance Program, a three-year intervention program for women who abuse alcohol and drugs during an index pregnancy. Pretest-posttest comparison was made across three sites: the original demonstration (1991-1995), and the Seattle and Tacoma replications (1996-2003). In the original demonstration, the client group performed significantly better than controls. Compared to the original demonstration, outcomes at replication sites were maintained (for regular use of contraception and use of reliable method; and number of subsequent deliveries), or improved (for alcohol/drug treatment completed; alcohol/ drug abstinence; subsequent delivery unexposed to alcohol/drugs). Improved outcomes at replication sites are not attributable to enrolling lower-risk women. Public policies and programs initiated over the study period may have had a positive effect on outcomes. Study findings suggest that this community-based intervention model is effective over time and across venues.