Criteria for participation in the study were pregnancy, the injection of any drug within the previous 6 months, consent for researchers to be informed of HIV status, literacy in English, and residence in metropolitan Sydney. In addition to the collection of basic information about demographic characteristics and duration since last injecting and sexual risk occasions, the HIV Risk-Taking Behavior Scale was administered to all subjects. The cognitive-behavioral relapse prevention intervention administered to the experimental group focused on the acquisition of skills aimed at helping prevent relapse to needle sharing and unsafe sex. Both experimental and control groups received counseling and advice about HIV-risk-taking behaviors normally available from methadone maintenance treatment program staff. Findings revealed there was no change in drug use in either group after the intervention. At 9-month follow-up, however, the experimental group had significantly reduced some HIV risk taking behaviors, in particular, injecting risk behaviors. The experimental group reduced the needle risk associated with typical use (drug use in the month before interview) and binge use (drug use in the month nominated as the heaviest month of drug use in the previous 6 months). The intervention had no effect on sexual risk behaviors. The finding of reduced injecting risk behavior following the six-session intervention program suggests such an intervention may benefit individuals who persist with injecting risk behaviors despite methadone maintenance treatment and the availability of sterile injection equipment. 32 references, 2 tables, and 4 figures