The role of ventilation in preventing tuberculosis (TB) transmission has been widely proposed in infection control guidance. However, conclusive evidence is lacking. Modelling suggested the threshold of ventilation rate to reduce effective reproductive ratio (ratio between new secondary infectious cases and source cases) of TB to below 1 is corresponding to a CO level of 1,000 parts per million (ppm). Here, we measured the effect of improving ventilation rate on a TB outbreak involving 27 TB cases and 1,665 contacts in underventilated university buildings. Ventilation engineering decreased the maximum CO levels from 3,204 ± 50 ppm to 591-603 ppm. Thereafter, the secondary attack rate of new contacts in university dropped to zero (mean follow-up duration: 5.9 years). Exposure to source TB cases under CO >1,000 ppm indoor environment was a significant risk factor for contacts to become new infectious TB cases (P<0.001). After adjusting for effects of contact investigation and latent TB infection treatment, improving ventilation rate to levels with CO <1,000 ppm was independently associated with a 97% decrease (95% CI: 50%-99.9%) in the incidence of TB among contacts. These results show that maintaining adequate indoor ventilation could be a highly effective strategy for controlling TB outbreaks.