OBJECTIVE: Medicaid expenditures for antipsychotic medications have risen rapidly, from under $1.0 billion in 1995 to over $5.5 billion in 2005. In response, at least ten states have implemented prior-authorization programs that restrict access to particular second-generation antipsychotic agents (aripiprazole and olanzapine). Twenty-two states restrict particular dosing forms (injections). This study examined the impact of such restrictions. METHODS: The authors used interrupted time-series analysis of quarterly state-level drug utilization data to examine the impact of prior authorization for particular agents in West Virginia and Texas. Changes in market share of nonpreferred medications and total pharmacy costs were compared with changes in states without similar prior-authorization requirements. RESULTS: The West Virginia policy led to an immediate 3.5% reduction in market share level (p<.01) and a 1.3% decrease in trend per quarter in market share (p<.001) for nonpreferred antipsychotics, leading to a 13.9% reduction after two years. In Texas, prior authorization reduced the market share level of nonpreferred agents by 2.6% (p=.055). However, prior authorization did not lead to a significant decrease in pharmacy reimbursements in either state. CONCLUSIONS: Current prior-authorization policies for second-generation antipsychotics do not appear to reduce pharmacy reimbursement, probably because alternative medications are costly. These findings suggest that any cost savings from prior-authorization policies would accrue largely through supplemental rebate agreements with manufacturers, which are likely reduced by the transfer of dually eligible Medicaid enrollees to Medicare Part D plans. Further evaluation of the clinical consequences resulting from such policies is urgently needed to determine whether the minimal cost savings outweigh the potential clinical risks.