Lung cancer risk among exsmokers according to years since cessation of smoking was assessed by means of a case-control study. The case series consisted of 1,052 lung cancer patients who were newly diagnosed and admitted to eight hospitals in Osaka in 1986-88. Smoking histories were compared with those of 1,111 controls admitted to the same hospitals during the same period without any diagnosis of smoking-related disease. The odds ratio of lung cancer for exsmokers compared to current smokers was estimated to be 0.90, 0.50, 0.51, 0.59, 0.48 and 0.29, for 1-4, 5-9, 10-14, 15-19, 20-24 and greater than or equal to 25 years after cessation of smoking, respectively. Risk reduction appeared to be greater for those who smoked less than the 1200 cigarette index, compared to those who smoked more. In classification according to histologic type, small cell and large cell carcinoma showed a rapid decrease compared to adenocarcinoma, while squamous cell carcinoma showed an intermediate pattern. Quantitative estimates for reduction of lung cancer risk among exsmokers can be used for projecting lung cancer incidence in the future, by assuming future trends of smoking prevalence, as well as for health education among individual smokers.