OBJECTIVE: To quantify the effect of diabetes-related hypoglycemia on direct health care costs and utilization. METHODS: The sample consisted of 9,162 privately insured employees that had at least 12 months of continuous enrollment between January 1, 1999 and December 31, 2001, and two or more prescriptions for diabetes-related medications within the same therapeutic class (insulin, sulfonylurea, or other oral diabetes-related medication) from Medstat's Health Productivity and Management database. Patients were grouped into two cohorts: hypoglycemia or non-hypoglycemia depending on whether they incurred a diagnosis of hypoglycemia (ICD-9-CM 250.8, 251.1, 251.2). The sample was further sub-set to those with claims for insulin (N = 2,664) - 64% of the original sample. Data were gathered starting from the first evidence of a diabetes-related medication. RESULTS: The incidence of hypoglycemia did not vary significantly by age, insurance type, or region of the country. However, the hypoglycemia group had evidence of significantly increased comorbidity including renal disease (14% vs. 6%, p <0.01), ophthalmic findings (42% vs. 30%, p <0.01), and neuropathy (30% vs. 15%, p <0.01), as well as a higher mean Charlson Comorbidity Index (1.20 vs. 0.75, p <0.01). Hypoglycemia patients incurred on average twice as many hospitalizations (1.0 vs. 0.5, p <0.01) and emergency room visits (0.8 vs. 0.4, p <0.01) annually compared to insulin-users without hypoglycemia. Forty percent (40%) of the excess hospitalizations and 53% of the excess emergency room visits were associated with a hypoglycemia diagnosis. Hypoglycemia patients on insulin incurred twice the amount of overall health care expenditures ($28,049 vs. $14,019, p <0.01) and three times the amount of diabetes-related health care expenditures ($14,590 vs. $5,236, p <0.01) compared to non-hypoglycemia insulin users. CONCLUSION: Hypoglycemia contributed significantly to medical care utilization and health care expenditures among patients using insulin. Analysis of the indirect costs among this population as well as investigation into hypoglycemia among diabetes patients on oral agents warrants further investigation.