Objectives Household air pollution (HAP) is a risk factor for respiratory disease, however has yet to be definitively associated with tuberculosis (TB). We aimed to assess the association between HAP and TB.

Methods A matched case–control study was conducted among adult women and children patients with TB and healthy controls matched on geography, age and sex. HAP was assessed using questionnaires for pollution sources and 24-hour household concentrations of particulate matter <2.5 μm in diameter (PM_2.5).

Results In total, 192 individuals in 96 matched pairs were included. The median 24-hour time-weighted average PM_2.5 was nearly seven times higher than the WHO’s recommendation of 25 µg/m³, and did not vary between controls (179 µg/m³; IQR: 113–292) and cases (median 157 µg/m³; 95% CI 93 to 279; p=0.57). Reported use of wood fuel was not associated with TB (OR 2.32; 95% CI 0.65 to 24.20) and kerosene was significantly associated with TB (OR 5.49, 95% CI 1.24 to 24.20) in adjusted analysis. Household PM_2.5 was not associated with TB in univariate or adjusted analysis. Controlling for PM_2.5 concentration, kerosene was not significantly associated with TB, but effect sizes ranged from OR 4.30 (95% CI 0.78 to 30.86; p=0.09) to OR 5.49 (0.82 to 36.75; p=0.08).

Conclusions Use of kerosene cooking fuel is positively associated with TB in analysis using reported sources of exposure. Ubiquitously high levels of particulates limited detection of a difference in household PM_2.5 between cases and controls.