OBJECTIVES To estimate future scenarios of utilization of knee arthroplasty (KA) revision in the Spanish National Health System at the short and long term and its impact on primary KA utilization. METHODS A discrete event simulation model was built to represent the utilization of KA for 20 years (2011–2031) in the Spanish National Health System, especially the burden of KA revision according to different scenarios of utilization and protheses survival. Data on KA utilization from 1997 to 2011 was obtained from the Spanish Minimum Data Set. Three scenarios of future utilization of primary KA were estimated: 1) fixed number since 2011; 2) fixed age and sex adjusted rates since 2011; and 3) projection using a linear regression model. These three scenarios were combined with two prostheses survival functions L) from a study including primary KA from 1995 to 2000; and H) from the Catalan Registry of Arthroplasty, including primary KA from 2005 to 2013. The model was programmed using ARENA. The simulation results were analyzed at the short (2015) and long-term (2030). RESULTS Variations in the number of revisions depended on both the primary utilization rate and the survival function applied, ranging from 8.3% to 31.6% increase at the short-term and from 38.3% to 176.9% at the long term, percentages corresponding to the combinations of scenario 1 (low primary utilization rate) and survival function H (better survival) versus scenario 3 (high primary utilization rate) and survival function L (worse survival), respectively. The prediction of increase on overall surgeries ranged from 0.1% to 22.3% at the short-term and from 3.7% to 98.2% at the long-term. CONCLUSIONS Projections of the burden of knee arthroplasty provide a quantitative basis for future policy decisions relating to concentration of high complexity procedures, the number of orthopaedic surgeons required to perform these procedures and the number of resources needed.