Tuberculosis (TB) is the leading cause of death among adults living with HIV in sub-Saharan Africa. Recent randomized clinical trial data has shown that earlier ART initiation (within two weeks for CD4 <50/µl and within two months for CD4 >50/µl) reduces mortality by 40 to 50%. There is a large gap, however, in the translation of this evidence into real world practice. In collaboration with the Center for Infectious Disease Research in Zambia (CIDRZ), we propose a mixed methods approach to assess this gap in the context of Zambia, a country with an HIV-TB incidence of 251/100,000 persons. First, we propose to use quantitative epidemiologic methods to estimate the timing of ART initiation among ART-naive patients with TB as well as the effect, through application of marginal structural models, of that timing on survival. The anticipated large sample size in this study (n=5,000) will be achieved through leveraging a parent study, recently funded by the Gates Foundation to evaluate treatment outcomes in the national HIV program. Using this study platform, we will collect additional specific measurements related to TB from clinical record review. Second, we propose qualitative interviews with 45 health professionals at all tiers of the Zambian health system (front-line healthcare workers, facility managers, and regional and national policy makers) to understand barriers to greater uptake of evidence-based early ART initiation. We will use Michie's Behavioral Wheel framework (an overarching theory to assess the behavior of health care professionals through the lens of three factors: capabilities, opportunities and motivation) to structure the interviews. The study will be conducted over a period of 12 months. Accomplishment of these aims will provide valuable scientific data and strongly position me to build a K23 application to further describe and evaluate the uptake of evidence-based practice in HIV/AIDS care in resource-limited settings.