OBJECTIVE: Publicly funded Medicaid prescription drug expenditures continue to escalate at rates far greater than those of private insurance programs. Cost containment policies like formularies or PDLs have been implemented widely to stem such growth. While a payer perspective is justified in evaluating the budgetary and program impact of formularies in private payer programs, a societal perspective is more appropriate in public funded programs like Medicaid. Objective of this study was to review published studies that evaluate the impact of Medicaid formularies, and to assess the perspective that was adopted for such evaluations. METHODS: Searches were performed in several scientific and business databases along with other relevant health policy journals and websites. The search strategy involved an incremental focus using a tiered approach, narrowing search returns down to the specific inclusion criteria with multiple combinations of search terms. Selected articles were reviewed to assess the perspective adopted based on key measurement parameters used in evaluation of the impact. RESULTS: Twenty studies were selected for the review. All studies used a pre-post or a time-series research design. Three studies utilized pooled multi-state data, two studies included a second state's Medicaid population as a control group, and the remaining were studies of single state Medicaid populations. All studies evaluated direct impacts of a formulary strategy on drug and program costs. Three studies additionally investigated broader second order effects on physician, outpatient and hospital utilization. None of the evaluations included estimates of indirect costs related to the positive or negative impact of formularies on morbidity, mortality and consequent productivity of Medicaid beneficiaries. Therefore, the true impact of formularies in these evaluations remain unclear. CONCLUSION: Findings of the literature review of Medicaid formulary evaluations suggest that such a societal perspective has not been widely adopted, and can be a valuable option for future evaluations.