OBJECTIVE: Budget impact analysis (BIA) evaluates financial impacts of new technologies; it provides valuable information to decision-makers with a budget concern. This study proposes a dynamic model to incorporate variations in patients mix and drug prices over time in BIA. METHODS: The dynamic model is an inhomogeneous Markov Chain model. It contains three Markov states categorized by whether a patient's illness was treated with a generic drug, an existing brand-name drug, or the new brand-name drug. At each cycle, the model modifies the patient cohort by accounting for newly diagnosed incident cases and exiting cases due to cure or death. Also considered is a possible difference in the preference of treatment selection between the current and newly diagnosed patients. We conducted BIA on a simulated data using the Bayesian approach and presented the results in a probabilistic plot similar to the cost-effectiveness acceptability curve. A case study comparing the budget impact of including versus excluding a new drug in a health plan was used to demonstrate our method. The case study assumes the perspective of a payer and a time frame of five years. RESULTS: Results based on the simulated data showed that adding the new drug to the plan is associated with a budget increase in the short run but would reduce the budget in the long run. The probability that including the new drug would increase in the budget by 10% is 9%, 26% in a one- and two-year study timeframe, and it becomes cost neutral in the five-year timeframe. CONCLUSIONS: The proposed model provides a powerful framework to examine the time-varying parameters in BIA and generates estimates that better reflect health care market in the real world.