After excluding participants with evidence of preexisting serious chronic disease at exam 1 or missing values on the predictor, outcome or covariate variables, there were 7802 participants with 861 new verified diagnoses of diabetes during 98,543 person years of follow up. Mean follow-up time was 13.3 years for censored individuals. The number of participants who provided an updated chocolate consumption variable at exam 3 was 6614. More frequent chocolate consumption was associated with lower risk of diabetes, up to 2-6 1 oz servings per week (Table 2). However, there was no significant association for 1 serving per day.