Current estimates indicate that about a third of the world’s population is infected with Mycobacterium Tuberculosis (TB). 8.6 million individuals actually develop the disease annually, including 1.1 million cases among people living with human immunodeficiency virus (HIV). 1.3 million people die annually from TB disease, including 320,000 HIV-positive individuals. In order to decrease the global TB burden, tuberculosis control programs have focused on disease prevention and transmission control simultaneously. Such programs are threatened by the spread of drug-resistant TB strains (strains that are resistant to at least one first-line treatment). Despite recent advances in the development of rapid tests designed to identify the genetic signatures of Mycobacteria tuberculosis (MTB) and its resistance conferring mutations, major gaps still exist. Not only has drug-resistant TB emerged in many parts of the world, strains of MTB are also becoming resistant to an increasing number of drugs. At the extreme are reports of “totally drug-resistant” strains in several countries. Early diagnosis and treatment of drug-resistant TB is essential both to improve patient outcomes and to interrupt further transmission of Drug-resistant (DR) strains. However, current tools either target only a limited number of drugs or have a low sensitivity. These problems result from both technical limitations of the tests and from a “discovery” gap -- the complete spectrum of mutations that cause resistance to these agents has not yet been identified. Leveraging an existing archive of TB strains that have undergone drug sensitivity testing and target sequencing, we perform a genome-wide association study (GWAS) on TB strains for which resistance phenotypes are unexplained by mutations in the known or suspected resistance genes.