An educational outreach programme promoting opt-out rapid HIV testing of new practice registrants led to increased rates of HIV diagnosis but there was no significant difference in the proportion having low CD4 counts at diagnosis. The study was based in Hackney, an area of London with a diagnosed HIV prevalence of 8 per 1,000. Practices were randomised into two groups. The intervention group received a practice-based outreach educational programme with follow-up training, integrated rapid HIV testing into the registration health check, provided free rapid HIV tests and were paid £10 per test completed.