ABSTRACT Every year 10 million people fall ill with tuberculosis (TB) and 1.5 million people die from TB – making it the leading cause of death from infectious disease – but TB can be difficult to diagnose. Extended M. tuberculosis (Mtb) cultures are still often used for diagnosis. PCR-based assays (e.g. Xpert MTB/RIF) that detect TB DNA can provide more rapid results, but require special equipment and, like Mtb culture, exhibit reduced sensitivity when employed to analyze sputum from individuals with extrapulmonary TB or compromised immune systems. Blood-based TB assays should detect all forms of TB, but current tests analyze the immune response to Mtb antigens and cannot distinguish active and latent TB. Sensitive detection of Mtb-derived cell-free DNA (cfDNA) in the circulation, however, represents a potential new means for enhanced TB diagnosis. Circulating cfDNA is rapidly degraded after its release and disease-associated cfDNA levels in blood can rapidly change in response to pathologic changes and physiologic responses. This short half-life can enable “real-time” cfDNA analyses required for accurate evaluation of the current status of active Mtb infections and rapid granular evaluation of Mtb treatment responses. However, the limit of detection (LoD) of current PCR-based assays are not sufficient to reliably detect Mtb cfDNA in blood. We have established a CRISPR-Cas12a-based detection system that can ultra-sensitively detect trace amount of SARS-CoV-2 RNA in blood to diagnose COVID-19, predict disease severity, and evaluate infection resolution. We have adapted this approach to develop a blood-based multiplexed CRISPR-mediated TB diagnosis (CRISPR-TBD) assay that can detect circulating Mtb cfDNA, including SNPs responsible for drug-resistant TB. Our preliminary data from longitudinal serum samples of patients undergoing TB treatment provide strong proof-of-principle evidence for the clinical utility of this platform. We have adapted this approach to a paper strip-based point of care (POC) CRISPR-TBD detection platform suitable for use in resource-limited regions with high TB burden, without decreasing assay sensitivity. We now propose to: 1) systematically optimize all CRISPR-TBD paper strip assay steps to improve quantitative detection of Mtb-cfDNA in POC settings; 2) evaluate the performance of this POC assay to diagnose pulmonary and extrapulmonary TB and to identify drug-sensitive and -resistant TB cases; 3) quantify Mtb-cfDNA changes in serum during TB treatment as a measure of treatment efficacy or failure and for early detection of nascent drug resistance; and 4) in-field validate the diagnostic performance of this POC assay in an independent TB patient cohort when performed in a clinical laboratory in a high endemic TB region and evaluate in parallel the performance our predictive Mtb-cfDNA model for TB treatment.