285 Background: Health insurance claims data are increasingly used for estimating individual health outcomes. However, it is challenging to obtain population-based estimates because of difficulties obtaining and reconciling data from all insurance providers. With incomplete insurance data, methods are needed to estimate the entire insured population. Using multi-payer claims we demonstrate and validate a synthetic method to obtain county level morbidity estimates. METHODS We used data from the Integrated Cancer Information and Surveillance System (ICISS) data resource at the University of North Carolina (NC) at Chapel Hill. ICISS data represent a linked data resource comprised of beneficiaries enrolled in federal as well as private insurance plans and rich ecologic data; and represent 5.5 million unique individuals, about 55% of the NC population. Using specific ICD-9 diagnosis codes, claims data from 2008 were compared to state department of public health data. We computed county level hospitalization rates by summing data from three sub-groups: age 65 and older in the 100% Medicare sample, age younger than 65 in the 100% Medicaid sample, and age younger than 65 represented in the private payer data. For the privately insured population, we used census data to obtain estimates of the entire privately insured population and used the hospitalization rate from beneficiaries in ICISS data to estimate a numerator for the synthetic sample of privately insured beneficiaries. We used State Inpatient Data (SID) from NC County Data Book to validate our method. RESULTS Overall, our synthetic approach showed moderate to high validity with a Pearson correlation coefficient 0.77 for heart disease and 0.93 for flu or pneumonia. Our hospitalization estimates were slightly lower than the data from SID, because SID data include uninsured individuals and multiple hospitalizations for individuals. CONCLUSIONS Our synthetic method can be useful in estimating population-based health outcomes using linked insurance claims data.