Background The effect of ART used before aids diagnosis has been the object of studies. The objectives were to estimate AIDS incidence rates, median AIDS-free-time and to investigate predictor factors for progression to AIDS.

Methods Retrospective cohort study, encompassing 1879 adult patients of the HIV São Paulo Cohort, during 1988–2005. The Kaplan-Meier methods, the Cox proportional hazard model and HRs estimates were used.

Results 981 patients progressed to aids. AIDS incidence rates were 11.6 and 7.1 person-years in the 1988–1996 and 1997–2003 periods, respectively. The median time of progression from HIV infection to AIDS without treatment was 53.7 months; with ART without HAART, 90.0 months; and with HAART, over 50% of patients followed did not progress to AIDS until 108 months. Independent prognostic factors for AIDS-free-time were: treatment with ART without HAART (HR 2.1; 95% CI 1.6 to 2.8), no treatment regimen (HR 3.0; 95% CI 2.5 to 3.6); age at HIV infection diagnosis between 30 and 49 years (HR 1.2; 95% CI 1.1 to 1.3), age over 50 years (HR 2.9; 95% CI 2.3 to 5.2); black race/colour (HR 1.4; 95% CI 1.1 to 1.7); MSM (HR 1.4; 95% CI 1.1 to 1.6) and IDU (HR 1.7; 95% CI 1.3 to 2.2) exposure categories; up to 8 years of schooling (HR 1.3; 95% CI 1.1 to 1.5) and no schooling (HR 2.0; 95% CI 1.4 to 5.6); and CD4 count between 350 and 500 cells/mm³ (HR 1.6; 95% CI 1.3 to 1.9).

Conclusions Increased AIDS-free-time was observed, with HAART. Decrease in the incidence rates were observed, Predictor factors to AIDS were treatment, age, race/colour, transmission categories, schooling and CD4 count.