Hospital readmissions at Memorial Sloan Kettering Cancer Center and elsewhere are common and expensive. Close to 30 percent of discharges result in readmission within 30 days. The Memorial Sloan Kettering Institutional Database (IDB) provides an invaluable resource for exploring correlates of patients’ likelihood of readmission, based on readily accessible data. Possible factors include patient demographics, disease category, treating physician and service, reason for admission, type of discharge, and prior admission history. Howard Thaler, PhD, will present some initial findings and discuss challenges in data management and analysis, statistical modeling, and interpretation of results.