Although widely used, the tuberculin skin test (TST) has limited diagnostic accuracy for Mycobacterium tuberculosis (M. tb) infection. TST sensitivity is particularly limited in children with compromised immune status due to HIV (Human Immunodeficiency Virus) infection, young age and/or severe malnutrition. Very limited data exists regarding the diagnostic utility of Interferon Gamma Release Assays (IGRAs) to detect M. tb infection in children. The proposed study will address major gaps in current knowledge regarding the diagnosis of M. tb infection in HIV-infected children and may be a powerful tool to spur policy change that would dramatically improve health in children in countries endemic for tuberculosis (TB) and HIV. The primary objectives of the study will be to assess the agreement between the TST and IGRAs in HIV-infected and uninfected children less than or equal to age 5 and to assess the performance of TST and IGRAs across a standard gradient of M. tb exposure in HIV-infected and uninfected children less than or equal to age 5. The secondary objectives of the study will be to measure the impact of potential interaction and confounding variables on the TST and IGRAs in HIV-infected and uninfected children less than or equal to age 5 and to identify factors that modify children's response to the TST and IGRAs over time. This study is a prospective community-based study in a setting highly endemic for HIV and TB, the agreement between the TST and IGRAs will be assessed in 800 children, ages 3 months to 5 years, and the influence of M. tb exposure, HIV status, cellular immunity, nutritional status and age evaluated. Subjects will include HIV-infected (N=300) and uninfected (N=500) children with variable degrees of M. tb exposure. The inclusion of children with no known household TB contact will result in a measure of background community M. tb exposure in a setting with high annual risk of M. tb infection (ARI). Active contact investigation of children exposed to newly diagnosed adult TB index cases in the household allows for the creation of a standardized M. tb exposure gradient to serve as a surrogate measure of M. tb infection in the absence of an existing gold standard. After determination of household membership, researchers will administer a standard questionnaire, assess M. tb exposure, TB disease and HIV status, collect approximately 6-8 ml of blood for IGRAs, place a TST (after phlebotomy) and read the TST at 48-72 hours. In sub-groups of children enrolled during the first 18 months, the TST, IGRAs and measures of cellular immunity will be repeated at 3, 6, 12 and 24-months post enrollment. This study has been designed to meet guidelines for complete and accurate reporting of studies on diagnostic accuracy. There will be one study site (The Desmond Tutu TB Center, Stellenbosch University, Tygerberg, Cape Town, South Africa). Study enrollment will occur over 3 years. The expected duration of subject participation will range from 2 days to 24 months. The primary (dependent) outcome is M. tb infection as defined by TST, T. Spot Tb, and Quantiferon-Gold In-tube (QTF) completed at baseline. The secondary (dependent) outcome is M. tb infection as defined by TST, T. Spot Tb, and QTF completed at 3, 6, 12, and 24 months. The primary independent outcome is M. tb infection as defined by M. tb contact score. The secondary independent outcome is HIV status, age, nutritional status, and cellular immunity (CD4 count).