Background: Three-drug antiretroviral therapy (ART-3) has reduced complications and improved survival in HIV+ patients. The Chilean Public Health System began dual therapy (ART-2) in 1997, covering approximately 40% of patients in need. Aim: To report the results of a follow-up of patients with and without access to ART in a Chilean public hospital. Patient and Methods: Prospective follow-up of patients with ART-2 and 3 (cases) and patients with no access to ART (controls). All patients needed ART but it was available to a minority of them. Selection for ART was at random. Follow-up was between 6-24 months (11/96 to 3/99). Basal and periodic clinical and laboratory parameters were determined. Mortality and occurrence of new AIDS-defining events (ADE) were compared statistically using chi square. Results: One hundred and fifty cases (106 ART2, 28 ART3 and 16 ART2 expanded to ART3) and 166 controls were studied. Basal parameters were similar except prior ART (32.7 and 18.7% in cases and controls respectively). Close to 1/3 patients had AIDS. Cases had a mean follow up of 527 days; controls, 478. Six cases (4%) (5 in ART-2) and 17 controls (10%) died. Mortality x 100/pts/yr was 2.7 in cases and 7.7 in controls, p <0.05. ADE per 100/pts/yr was 21 in cases (24.4 in ART2, 15.1 in TAR3) and 54.5 in controls, p <0.05. Cases had a reduction of: esophageal candidiasis (84%), tuberculosis (75%), cryptococcosis and toxoplasmosis (66%), P carinii pneumonia (55%) and bacterial pneumonia (46%) and they had fewer hospitalizations (73%). Late assessment: 70 of 101 ART-2 patients had changed to ART-3 (1 death); 22 of 101 kept original ART-2 (12 dead, 10 alive), 39 of 43 ART-3 patients were alive and 1 died. Conclusions: Short-term ART-2 and 3 significantly reduced mortality (60% and 73%) ADE (65% and 76% respectively) and hospitalizations. Benefits of ART-2 were short lived. Resource-constrained countries cannot depend on weaker than standard ART for proper care of people with HIV disease. (Rev Méd Chile 2001; 129: 886-94)