OBJECTIVES: To evaluate the effect of the July, 2001 prescription limit policy change in the South Carolina Medicaid program on the utilization of health care services and their related costs for adult Medicaid recipients diagnosed with schizophrenia or bipolar disorder. METHODS: A retrospective cohort study design, identifying subjects with schizophrenia and bipolar disorder compared their utilization of health care services and associated costs 18-months before and after the policy change (July, 2001). Eligible patients were age 21 or older, had a qualifying diagnosis on a hospital or ambulatory claim, and a prescription medication for their diagnosis within 90 days (+/-) of their 1998 or 1999 enrollment date. Total health care cost and service utilization were estimated by ordinary least squares regression models and the results contrasted with panel regression methods due to the short time series. Predictor variables were demographics, inpatient hospitalization, and comorbidities. RESULTS: OLS and panel estimation show an increase in total cost and the number of ambulatory, hospital, prescription, and nursing home claims after the policy change. Panel estimation shows a positive monthly trend in the post period for all claims. CONCLUSIONS: The policy change resulting in an increase in average monthly patient prescription cost is associated with increases in total cost of care and overall health care utilizations.