The role of microbial factors in outcomes of tuberculosis treatment has not been well studied. We performed a case-control study to evaluate the association between a Beijing strain and tuberculosis treatment outcomes. Isolates from patients with culture-positive treatment failure (n = 8) or relapse (n = 54) were compared with isolates from randomly selected controls (n = 296) by using spoligotyping. Patients with Beijing strains had a higher risk for relapse (odds ratio [OR] 2.0, 95% confidence interval [CI] 1.0-4.0, p = 0.04) but not for treatment failure. Adjustment for factors previously associated with relapse had little effect on the association between Beijing strains and relapse. Beijing strains were strongly associated with relapse among Asian-Pacific Islanders (OR 11, 95% CI 1.1-108, p = 0.04). Active disease caused by a Beijing strain was associated with increased risk for relapse, particularly among Asian-Pacific Islanders.