Abstract Background Tuberculosis (TB) natural history remains poorly characterized, and new investigations are impossible as it would be unethical to follow up TB patients without treatment. Methods We considered the reports identified in a previous systematic review of studies from the prechemotherapy era, and extracted detailed data on mortality over time. We used a Bayesian framework to estimate the rates of TB-induced mortality and self-cure. A hierarchical model was employed to allow estimates to vary by cohort. Inference was performed separately for smear-positive TB (SP-TB) and smear-negative TB (SN-TB). Results We included 41 cohorts of SP-TB patients and 19 cohorts of pulmonary SN-TB patients in the analysis. The median estimates of the TB-specific mortality rates were 0.389 year−1 (95% credible interval [CrI], .335–.449) and 0.025 year−1 (95% CrI, .017–.035) for SP-TB and SN-TB patients, respectively. The estimates for self-recovery rates were 0.231 year−1 (95% CrI, .177–.288) and 0.130 year−1 (95% CrI, .073–.209) for SP-TB and SN-TB patients, respectively. These rates correspond to average durations of untreated TB of 1.57 years (95% CrI, 1.37–1.81) and 5.35 years (95% CrI, 3.42–8.23) for SP-TB and SN-TB, respectively, when assuming a non-TB-related mortality rate of 0.014 year−1 (ie, a 70-year life expectancy). Conclusions TB-specific mortality rates are around 15 times higher for SP-TB than for SN-TB patients. This difference was underestimated dramatically in previous TB modeling studies, raising concerns about the accuracy of the associated predictions. Despite being less infectious, SN-TB may be responsible for equivalent numbers of secondary infections as SP-TB due to its much longer duration.