OBJECTIVES:  Assess the cost-effectiveness of Sunitinib (SU) versus Pazopanib (PA) and Best Supportive Care (BSC) for the treatment of metastatic renal cell carcinoma (MRCC) from the perspective of the Chilean public healthcare system. METHODS:  A four health states Markov model was built: first and second line treatments, BSC and death. Expected costs were measured in Chilean pesos (1 USD = 654.07CLP$) and benefits in quality adjusted life years (QALYs) from aggregated data. Efficacy estimates were obtained from an indirect-treatment comparison analysis and costs were estimated from local sources. Utilities were obtained from the literature. A 10-year time horizon and 3% discount rate was considered for costs and outcomes. A probabilistic sensitivity analyses was performed to account for uncertainty. RESULTS:  The total expected costs of treating MRCC with SU or PA are higher than BSC with slight differences between SU and PA (US$45,786 and US$43,255 respectively). Similarly, the expected incremental health benefit is small favoring SU (0.03 QALYs). However, the uncertainty around this estimate is important leading to a non-negligible probability that it may actually favor either treatment alternative. The base case scenario (current market prices for all treatments) shows an average ICER of PA versus BSC of US$62,327/QALY and an ICER of SU versus PA of US$85,885/QALY. At a suggested threshold of 3xGDP per capita (US$55.040), the probability of cost-effectiveness of SU and PA was 25% and 32% respectively. This probability increases to 38% and 45% respectively when their price is reduced a 25%. The ICER was most sensitive to the OS efficacy relative to BSC, where evidence showed important bias due to cross-over. CONCLUSIONS: Due to the limited expected incremental QALYs and high second order uncertainty, it is reasonable to manage this decision problem as a cost minimization exercise which is determined by the price of each treatment alternative.