Of 468 diagnosis-related groups identified by the federal government for Medicaid reimbursement, 15 are related to obstetric hospital care. Each diagnosis-related group is considered a distinct group in which cases are homogeneous with respect to resource consumption. Because the diagnosis-related group system is based primarily on data from community and secondary care hospitals, it does not differentiate sufficiently among high-risk obstetric patients seen at tertiary care institutions, such as Florida's Regional Perinatal Intensive Care Centers. We developed an alternative scheme for diagnosis-related groups, called obstetric care groups, using the federal diagnosis-related groups as the model from which to depart. Data collected for 4192 women during a 2 1/2-year period indicate that obstetric care groups provide more homogeneous groups than diagnosis-related groups for our population of high-risk patients. The obstetric care groups differentiate between no complications, one complication, and two or more complications, while the diagnosis-related groups differentiate only between no complications and one or more complications. Also, complications for obstetric care groups are based on only 19 diagnoses that contribute significantly to resource consumption, while the list of possible complications exceeds 200 for diagnosis-related groups. Although the obstetric care group classification system is simpler than that for diagnosis-related groups, it results in a more accurate reimbursement of hospitalization charges for high-risk obstetric care.