OBJECTIVE: The incidence of some cancers has been reported to be higher in diabetic patients than in the general population. We estimated the incidence of lung cancer in diabetic patients and investigated the hypothesis that the rate of lung cancer is different in diabetic compared with nondiabetic patients. RESEARCH DESIGN AND METHODS: Diabetic patients and age-, sex-, and general practice-matched nondiabetic control subjects were identified from U.K. computerized general practice records (General Practice Research Database), and these records searched for any incident lung cancer, demographic details, and smoking status. Primary lung cancer incidence was calculated and rates compared between diabetic patients and nondiabetic control subjects using multivariate Cox regression, adjusting for age, sex, and smoking. The comparison was repeated for incident diabetic patients followed from diagnosis and after stratifying by diabetic treatment. RESULTS: The incidence of primary lung cancer in all 66,848 diabetic patients was 1.63 per 1,000 patient-years (95% CI 1.48-1.79) and 2.05 per 1,000 patient-years (1.76-2.38) among diabetic patients followed from diagnosis. When compared with nondiabetic control subjects, the hazard ratio was 0.88 (0.79-0.97) for all diabetic patients and 1.12 (0.95-1.34) for those followed from diagnosis. When observation was truncated to allow for shorter life expectancy, the hazard ratio for the total cohort was 0.98 (0.84-1.13), and no association was found with any treatment group. CONCLUSIONS: No increased risk of lung cancer in diabetes was found. We hypothesize that the lower incidence may be partly due to shorter life expectancy.