Involuntary smoking contains human carcinogens. Exposure prevalence among adults is on the order of 40%. A meta-analysis of epidemiologic studies on lung cancer and exposure to involuntary smoking from the spouse included 51 studies. The overall relative risk (RR), based on 7369 cases of lung cancer, was 1.25 [95% confidence interval (95% CI) 1.15-1.37]. No evidence existed of an RR difference between the two genders, and study design had no influence on the results. The summary RR was lower for adenocarcinoma than for other histological types. In the largest studies cumulative exposure suggested a dose-response relationship with a unit risk of similar magnitude. The summary RR was 1.17 (95% CI 1.04-1.32) for workplace exposure. Several sources of bias may lead to both overestimation and underestimation of true association, and the most plausible interpretation favors a causal association. Even if excess risk from exposure to involuntary smoking is small, its large prevalence makes it an important environmental carcinogen.