The study examined the association of diabetes mellitus (DM) history with total and common site-specific cancersusing a large cohort of 23,378 men and 33,503 women, extracted from 127,477 healthy participants of the JACCStudy who were aged 40-79 years and living in 24 municipalities in Japan. At enrolment during 1988-90, eachsubject completed a self-administered questionnaire including items for age, sex, body mass index (BMI), smoking,drinking, past history of DM and cancer. Adjusting for age, BMI, smoking, and drinking in the Cox’s proportionalhazard model, incidence rate ratios (IRR) with 95% confidence intervals (95%CIs) were estimated for both sexes.During the follow-up period, total cancers and site-specific cancers were identified. A history of DM was reported by7.5% of men and 4.6% of women. DM significantly increased the risk of liver cancer for both men (IRR=2.30;95%CI=1.47-3.59) and women (IRR=2.70; 95%CI=1.20-6.05). Significant increased and reduced risk due to DM formen were also found for non-Hodgkin lymphoma (IRR=2.77; 95%CI=1.04-7.38) and stomach cancer (IRR=0.67;95%CI=0.46-0.99) respectively. For females, a reduced risk of stomach cancer due to DM (IRR=0.49; 95%CI=0.23-1.04) was also revealed. Since a history of DM here demonstrated significant associations with some site-specificcancers, their relationships should be studied further in Japan for validation.