OBJECTIVES: Performing a 10-year horizon (2007/2016) cost-effectiveness analysis of the integrated Spanish Renal Replacement Therapy (RRT) program, from the National Health System perspective. METHODS: A deterministic Markov chain model, allowing for incident patients, was designed to predict transitions between four states (hemodialysis (HD), peritoneal dialysis (PD), renal transplant (TX), and death (D)) for the period 2007/2016, providing future incidence, prevalence, and survivability estimates. Transition and mortality probabilities were calibrated using data from the Spanish Nephrology Society (SEN) and the Spanish National Statistics Institute (INE) for the period 1996/2006. Information on direct and indirect costs was gathered from several sources, including OBLIKUE database, regional official bulletins, literature review, and expert judgment. Data on utilities were obtained from a proprietary database (FIS 96/1327 Project). A sensitivity analysis on the main parameters of the model was performed in order to ensure that results were consistent. RESULTS: For the 2007/2016 period, average patients per year (HD, PD, and TX) were, respectively: 29,981; 3,348; 26,301. Average costs per patient and year (in 2007 discounted Euros) were: 39,026; 26,971; and 13,085. Average QALYs per patient and year were: 0.51; 0.55; and 0.68. Therefore, average cost-effectiveness was: 75,927; 49,478; and 19,253. TX was the optimal treatment, regarding cost-effectiveness estimates. Consequently, it was considered as the gold standard. Incremental Cost-Effectiveness Ratios were: -103,233 (TX vs. HD) and -156,636 (TX vs. PD). Net Health Profits were: 1.03 (TX vs. HD) and 0.60 (TX vs. PD). Consistency of these findings was supported by the sensitivity analysis. CONCLUSIONS: As expected, TX is the most convenient treatment, as regards cost-effectiveness estimates. PD is also preferred to HD. National Health System authorities should, therefore, favor PD over HD. If that is the case, a Budget Impact Analysis would help to assess to which extent such approach would be worthy.