The authors are supervising a multicenter study on a large cohort of intravenous heroin users for whom the date of seroconversion (the availability of a negative test preceding the confirmed positive one was a criterion for enrollment) could be identified. The aim of the study was to evaluate the risk of AIDS in HIV seroconverters who are drug injectors and to identify cofactors associated with the disease progression. The authors enrolled 201 newly seropositive individuals. Follow-up time was distributed over a 4-y period, because in some of the enrollees the test was performed retrospectively. The actuarial incidence of AIDS in the cohort, estimated by the Kaplan-Meier survival technique, was 1.5% in the first year after seroconversion but 18% in those followed for 4 y after seroconversion.