We investigated the prophylactic effect of dapsone on the development of Mycobacterium avium-intracellulare (MAI) in a retrospective analysis of an observational database of the Multicenter AIDS Cohort Study. Dapsone generally is not a first-line prophylactic agent for Pneumocystis carinii pneumonia (PCP), the time at risk for disease development increases with the length of survival time with low CD4 lymphocyte counts the subjects were not randomized to their therapies; for these reasons, we used time-dependent multivariate Cox models to control for selection, disease stage survival stage. We included 1035 subjects in the model, of whom 178 (17.2%) developed MAI and 216 (20.9%) used dapsone at some time. The mean length of follow-up was 3.28 ±2.1 years, mean length of time to development of MAI for those who had this disease was 2.38 ± 1.71 years mean time on dapsone therapy for those who used it was 1.17 ± 0.94 years. The following factors were all associated with a significantly increased risk for subsequent development of MAI: the location (city) of the subject, decreased CD4 and CD8 lymphocyte counts, presence of human immunodeficiency virus symptoms development of cytomegalovirus disease. The use of dapsone was associated with a nonsignificant 30% decrease (relative hazard = 0.7; 95% confidence interval = 0.6–1.37) in the risk of development of MAI after controlling for other factors in the time-dependent Cox model. Dapsone did not have a significant effect on development of MAI; however, given the wide confidence interval, its low cost the additional benefit of prophylaxis against PCP and toxoplasmosis, further investigation in combination with other agents is warranted.