OBJECTIVES: To assess the cost-effectiveness of leflunomide monotherapy versus cyclosporin monotherapy in metothrexate-resistant rheumatoid arthritis from the perspective of National Health Fund in Poland. METHODS: A decision analytic model, based on framework proposed by Kavanaugh was developed. A systematic review was conducted to collect clinical efficacy data. Indirect comparison with placebo as a common comparator was performed. Direct medical costs including costs of drugs, therapy monitoring, adverse reactions treatment, hospitalizations due to progression of arthritis were analyzed. Because of short, 6-months time horizon, discounting was not performed (values in Polish zloty (PLN): 1 Euro=3.8 PLN). Univariate sensitivity analysis was conducted for key clinical and cost variables. RESULTS: Leflunomide and cyclosporin based strategies resulted in 0.449 ACR20 and 0.323 ACR20 response, with corresponding costs of 2740 PLN and 2730 PLN, respectively (cost-effectiveness ratios were 6106 PLN/ACR20 and 8453 PLN/ACR20, respectively). Switching from cyclosporin to leflunomide based therapy resulted in additional ACR20 response gain for the price of 81 PLN. One-way sensitivity analysis revealed that clinical efficacy reduction of 25% for leflunomide and increase of 25% for cyclosporin resulted in incremental cost-effectiveness ratio rise to 3548 PLN/ACR20 and 6106 PLN/ACR20, respectively. CONCLUSION: In case of methotrexate resistance, leflunomide monotherapy is more cost-effective than cyclosporin monotherapy from the perspective of Polish National Health Fund.