OBJECTIVES: To develop a cost-effectiveness and a budget impact analysis of Rivaroxaban in the prevention of thromboembolic events in patients performing hip and knee arthroplasty in comparison with no treatment under the Brazilian private health care system perspective. METHODS: A decision tree analysis was developed for the first 90 days, considering the occurrence of Deep Venous Thrombosis, Pulmonary Embolism and thromboembolic events, followed by a Markov model, for Post Thrombotic Syndrome and Thrombotic Pulmonary Hypertension. The time horizon of the analysis was 5 year. The cycle duration was 1 year and the corresponding epidemiological and efficacy data were obtained from a critical appraisal of the scientific literature. Unit costs for drugs, procedures, materials and daily hospital were obtained from Kairos Magazine (Maximum price consumers 18%ICMS), Hierarchical Brazilian Classification of Medical Procedures (CBHPM 5th edition), Simpro Magazine Maximum price consumers 18% ICMS) and search UNIDAS 2008, respectively. A budget impact analysis was developed considering an increase of 10% per year in market share of Rivaroxaban. RESULTS: Total costs associated with Rivaroxaban and no treatment, considering the indication for knee arthroplasty, were BRL363 (US$214) and BRL1,040 (US$612), respectively. And considering the indication for hip arthroplasty, were BRL332 (US$195) and BRL462 (US$272), respectively. Rivaroxaban reduces the number of all thromboembolic events in 0.0793 and 0.0246, for knee and hip arthroplasty, respectively. Rivaroxaban treatment is more effective and cheaper than no treatment in both indications (dominant). The high cost associated with no treatment patient is due to the high number of events in this group. The budget impact analysis estimated an economy of BRL206,165 (US$121,274) and BRL104,351 (US$61,383) for knee and hip indication, respectively, in 5 years. CONCLUSIONS: By this pharmacoeconomic analysis, the treatment with Rivaroxaban, shown to reduce treatment costs and events compared with no treatment.