To date, nationally representative estimates of diabetes incidence have come from cross-sectional surveys, thus limiting the estimates to population proportions. The SUPREME-DM project combined longitudinal data from 11 HMOs to estimate actual diabetes incidence rates. Between 2005 and 2009, we identified 6,663,139 HMO members aged > 20 years who had > 2 years of membership eligibility without indication of diabetes. Members were followed from January 1, 2005 or their earliest date of eligibility after that date until they developed diabetes, died, disenrolled, or until December 31, 2009. New cases of diabetes over the 5-year period were identified using laboratory test results, pharmacy data (prescriptions for anti-diabetic medication), and inpatient and outpatient diagnosis codes. Anti-diabetic medication use during pregnancy was not included. We estimated incidence per 1,000 person-years for each participating HMO and then combined them using a membership-weighted average. We estimated incidence overall and by age/sex categories.Men had a 5-year diabetes incidence rate of 11.6/1000 person-years; the rate for women was 9.5/1000 person-years. For both sexes, incidence increased steadily across age groups 20-34, 35-44, 45-54, and 55-64 while it remained stable for age groups of 65-74, 75-84 and 85+ (see fi gure). Age/sex trends in incidence were nearly identical across all 11sites, but HMO's with higher minority populations had higher incidence rates across all age/sex groups. To our knowledge, these are the first estimates of diabetes incidence rates from a nationally-representative diverse insured cohort.