Drug plan decision makers need accurate financial impact projections for planning new drug policies. Projections should have minimal margins of error and be transparent and easy to communicate to stakeholders. OBJECTIVES: We explain how ad hoc methods typically used for financial impact projections are inadequate. METHODS: We describe a flexible tool for projecting the financial impact of drug policy changes based on historical dispensing data. The tool uses a random sample of a drug plan's beneficiaries to simulate the drug claim adjudication process under the proposed policy regulations. We explore the validity of the simulation tool using a recent example of a complex drug policy change in British Columbia (BC). Over 500 different policy options were simulated in the planning phase of the BC policy. Drug plan spending was projected for each option before the final policy was selected two months prior to the policy start. RESULTS: Predicted future total spending for the chosen policy option was within 1% of actual spending in the first 11 months ($555.8M and $560.0M, respectively). The average difference per week between actual and predicted amounts was 0.015% ($86,500, SD: $968,700). CONCLUSIONS: Such policy simulation can be applied to a wide range of health plans and policy changes.