Evidence on the spatial-temporal distribution of TB mortality across China and its relationship with long-term PM2.5 exposure is limited. We aimed to address significant gaps in our understanding of the spatial-temporal clustering patterns of tuberculosis mortality in China and provide evidence for its causal links with long-term PM2.5 exposure. Annual pulmonary tuberculosis mortality, PM2.5 concentrations, and socioeconomic factors for provinces in mainland China between 2004 and 2017 were obtained. Turning points in the temporal trend and spatial clustering patterns of pulmonary tuberculosis mortality were identified. A difference-in-differences causal inference approach was applied to estimate the long-term effect of PM2.5 exposure on the mortality. The average annual percent change of pulmonary tuberculosis mortality in China was -2.5% (95%CI: -5.6%, 0.7%), with an 11.1% annual increase in the Northwest since 2012 (P=0.029). The hot and cold spots were all located in northern China, where Xinjiang in the Northwest had the highest mortality across the study period. We found a significant association between long-term PM2.5 exposure and pulmonary tuberculosis mortality, with percent increases in risks (IR%) being 0.74% (95CI%, 0.04%, 1.45%). This association varied across multiple socioeconomic groups, with the highest IR% in provinces with lower level of latitude (IR%=0.83%, 95%CI: 0.01%, 1.65%), lower quartile of GDP (IR%=1.01%, 95%CI: 0.23%, 1.80%) or higher proportion (≥14%) of people > 65 years of age (IR%=1.24%, 95%CI: 0.44%, 2.04%). Comprehensive sensitivity analyses showed a robust adverse effect of long-term PM2.5 exposure on pulmonary tuberculosis mortality. Attention needs to be paid to the rising trend of pulmonary tuberculosis mortality in Northwest China. Our study provides the stable evidence to date of the causal association between long-term PM2.5 exposure and the risk of death from pulmonary tuberculosis, especially in low-altitude, underdeveloped, and aged provinces.