OBJECTIVES: Medical care costs for rhinitis are primarily driven by patient care-seeking behavior and physician prescribing patterns, which may evolve over time. Estimating a model of real-world rhinitis treatment from clinical trial data is not feasible due to short trial durations and protocol-driven care. Therefore, we used U.S. healthcare claims data to model rhinitis treatment patterns and estimate the budgetary impact of a novel rhinitis therapy. METHODS: We developed a three-year budgetary impact model of rhinitis using Markov-modeling techniques. Transitions between treatment regimens (monotherapy, dual-combination therapy, tri-combination therapy), treatment patterns, (therapy switching, add-on rates), and associated medical-care costs, were estimated from a large claims database, by identifying rhinitis patients and tracking changes in therapy over time. Budgetary impact of a novel treatment was assessed for three effectiveness scenarios, where the switching/add-on rates relative to fluticasone propionate, an existing rhinitis therapy, were 50% lower (Scenario 1), 25% lower (Scenario 2) and identical (Scenario 3). The novel treatment was assumed to be priced the same as fluticasone propionate and have a market share of 10%. RESULTS: The claims analysis found annual rates of treatment switching, add-on, and remaining on initial therapy ranging from 6-18%, 20-28%, and 62-72%, respectively, for currently existing rhinitis therapies. Annual rhinitis-related medical costs associated with each treatment pattern were $666, $657, and $558, respectively. In Scenario 1, the model predicted the per-patient-per-month (PPPM) budgetary impact for the novel treatment to be -$0.06, -$0.09, and -$0.11, in years 1-3, respectively. Scenarios 2 and 3 had corresponding PPPM results of $0.00, -$0.01, and -$0.01, and $0.05, $0.06, $0.08. CONCLUSION: Using claims data and Markov-modeling techniques, we found that budgetary impact can be materially affected by rates of treatment switching/add-on. Detailed, claims-based data are required for this type of analysis, given the real-world nature of treatment patterns and associated medical costs.