The overall objective of this proposal is to investigate the association between tobacco use and tuberculosis in HIV-infected males in Johannesburg, South Africa where rates of all three epidemics are extremely high. We will conduct a case-control study restricting all participants to HIV-infected men from Soweto. There will be two Case groups: 1) HIV-infected males with confirmed pulmonary tuberculosis and 2) HIV-infected males with extra-pulmonary TB and no evidence of pulmonary TB. The Control group will be HIV-infected males without TB. We will estimate nicotine exposure in all cases using hair samples, a biomarker that is relatively easy to collect and store and provides reliable exposure data of the previous month. We will also collect data for factors that may modify the association between tobacco and TB among HIV-infected patients including alcohol exposure and socioeconomic status.