In order to clarify the relation between cigarette smoking and lung cancer, a case-control study was conducted. The case series consisted of 1,376 lung cancer patients (1,082 males and 294 females) who were newly diagnosed and admitted to eight hospitals in Osaka during 1986-88. Smoking histories were compared with those of 2,230 controls (1,141 males and 1,089 females) admitted to the same hospitals during the same period without established smoking-related diseases. Odds ratios of current smoker versus nonsmoker were 18.1, 1.9, 21.4, and 3.8 for squamous, adeno, small, and large cell carcinoma, respectively, for males, and 9.7, 1.3, 12.1, 3.7, respectively, for females. Compared to the results from previous studies in Japan, the magnitude of the odds ratios for squamous and small cell carcinoma is approaching the level of Western Europe in the late 1970s. Population attributable risk of exsmokers has also been increasing to the level of Western Europe. Among male current smokers, smoking intensity, such as number of cigarettes per day or fraction smoked per cigarette, seemed to have a slightly greater influence on squamous cell carcinoma than adenocarcinoma, while factors associated with the spread of cigarette smoke, such as inhalation, seemed to have greater influence on adenocarcinoma. The difference in the distribution of these smoking characteristics between Japan and Western Europe could not fully explain the difference in lung cancer incidence and distribution of histologic types between the two areas.