Background A positive association between smoking and increased risk of tuberculosis disease is well documented for populations outside the USA. However, it is unclear whether smoking increases risk of tuberculosis in the USA, where both smoking prevalence and disease rates are much lower than in the countries where previous studies have been conducted.

Methods To explore the tuberculosis–smoking association in a more generalisable US population-based sample, we conducted a nested case–control study among members of Kaiser Permanente Northern California (KPNC). We identified all newly diagnosed cases of active pulmonary tuberculosis (PTB) disease between 1996 and 2010. Each of the 2380 cases were individually matched to two controls by age, gender and race/ethnicity. ORs and 95% CIs for the association between smoking status and PTB were calculated using conditional logistic regression adjusted for all matching factors.

Results Increased PTB risk was observed among ever-smokers (OR=1.35; 95% CI 1.19 to 1.53), as well as current (OR=1.26; 95% CI 1.08 to 1.48) and past (OR=1.43; 95% CI 1.23 to 1.67) smokers, compared with never-smokers. Increased intensity and duration of smoking were also positively associated with PTB risk.

Conclusions Our findings among a more generalisable US population support the hypothesis that smoking increases risk of PTB, underscoring the importance of tobacco cessation and prevention programmes in eliminating tuberculosis.