Studies on passive smoking have consistently shown a tendency toward an increased risk of breast cancer, while studies an active smoking have failed to demonstrate an association. This apparent contradiction may stem from not separating passive smokers from the unexposed when assessing the effect of active smoking. a population based case-control study was conducted in Geneva, Switzerland, between January 1992 and October 1993 to determine the relation of passive and active smoking to breast cancer when the referent unexposed category consisted of women unexposed to active and passive smoke. The 244 patients with breast cancer (cases) were compared with 1.032 women free of breast cancer (controls). The lifetime history of active and passive smoking was recorded year by year, between the age of 10 and the date of the interview. The adjusted odds of breast cancer for ever active smokers, compared with women unexposed to either passive or active smoke, were 2.2 (95% confidence interval (CI) 1.0-4.4) for an average lifetime consumption of 1-9 cigarettes per day, 2.7% (95% CI 1.4-5.4) for 10-19 cigarettes per day, and 4.6 (95% CI 2.2-9.7) for 30 or more cigarettes per day. Among passive smokers, the adjusted odds ratio was 3.2 (95% CI 1.6-6.3) for being exposed for the equivalent of 2 hours per day for 25 years. The odds ratios were adjusted for known or postulated risk factors of breast cancer, including alcohol and saturated fat intake. There was no evidence of strong selection, detection, or recall biases. Active and passive exposure to tobacco smoke may increase the risk of breast cancer. Additional studies are needed to decide whether the association is causal. Further elucidation of this relation would benefit not only the prevention of breast cancer but also the the prevention of other smoking-related diseases in women.