ABSTRACT M. tuberculosis (Mtb) infects one-third of the world’s population, causes TB disease in 10 million people per year, and is responsible for 1.6 million deaths annually. While effective treatments for TB have been available for over 50 years, a major barrier to its control has remained the inability to reliably diagnose disease in low-resource TB-endemic settings. Most clinical tests require an adequate sample of sputum (mucus from the lungs) which may be difficult to obtain in patients without a cough, in children, and in HIV-infected individuals. Urine offers an attractive method for point-of-care TB diagnostics since it is abundant, simple to collect, and collection does not involve exposure to infectious aerosols. Transrenal DNA (trDNA) present in the urine offers a specific target that can be amplified for high sensitivity. The development of a PCR-based test of TB from urine, and its development as a point-of-care test, could have enormous impact on TB diagnostics. Aim 1. To improve detection of Mtb-specific trDNA using ultrashort PCR and multiplexing. Hypothesis: Amplification of shorter fragments and/or targeting multiple unique fragments will enable detection of low- concentrations samples, increasing sensitivity while maintaining specificity. Aim 2. To determine sensitivity and specificity of improved Mtb-specific trDNA detection in urine samples from adults and children in a TB-endemic setting. Hypothesis: Mtb trDNA is present in detectable quantities in urine from both adults and children with pulmonary TB regardless of HIV status, and our hybridization sequence-capture method can increase clinical sensitivity compared to previous TB trDNA studies.