A new system for monitoring HIV voluntary counselling and testing (VCT) outcomes was established in 2007 at seven VCT clinics in Jakarta and Bali, Indonesia. Counsellors collected demographic and risk information from VCT clients. Logistic regression was used to determine factors associated with HIV infection. In 15 months, HIV prevalence in 5569 new clients without HIV symptoms was 63.3% in injecting drug users (IDUs) (n = 783), 7.7% in female sex workers (n = 1437), 31.6% among transgender people (n = 395), 9.3% in men who have sex with men (n = 268), 13.5% in clients of sex workers (n = 643), 21.1% in people with high-risk partners (n = 569) and 3.2% in other VCT clients (n = 822). Among IDUs, being older, tested though outreach, tested due to being 'at risk' and having injected for one or more years were independently associated with HIV infection. This network confirmed high HIV prevalence among IDUs and transgender people. HIV prevalence estimates were consistent with serosurveys.