PURPOSE: To assess the impact of an adolescent sexual health intervention on the use of health services by young people in Tanzania. METHODS: Twenty communities, including 39 health facilities, were randomly allocated to the intervention or comparison arm. Health workers from the intervention arm were trained in the provision of youth-friendly health services, as part of a package of interventions. Independent process evaluations were conducted in health facilities, and simulated patients visited clinics using sexual and reproductive health problem scenarios. The impact on health facility attendances were assessed in 1998 (baseline) and 1999-2001. Reported sexually transmitted infection (STI) symptoms and use of health services were evaluated in young people in the trial cohort. RESULTS: The mean monthly attendance for STI symptoms per health facility, per month was .5 for young males and 1.0 for young females at baseline. Attendance by young males was greater in the intervention communities in 1999-2000 after adjustment for baseline differences (p = .005), and this difference increased over time (p-trend = .022). The mean difference in attendance was however relatively modest, at 1.1 per month in 2001 after adjustment for baseline (95% CI: .5, 1.7). There was weaker evidence of an intervention effect on attendance by young women (p = .087). Few condoms were distributed, although a greater number were distributed in intervention facilities (p = .008). Generally, intervention health workers tended to be less judgmental and provided more comprehensive information. CONCLUSIONS: Training staff to provide more youth-friendly health services can increase the utilization of health services for suspected STIs by young people, especially among young men.