OBJECTIVES Knee osteoarthrosis is a multifactorial, progressive and incurable rheumatic ailment; most treatments look for a maximum recovery of mobility and functionality of the knee joint, with a minimum risk possibility. Due to its high cost and invasive character, gonarthrosis surgical treatment is reserved, according to the clinical practice guidance available in Mexico, for severe pain and joint functionality limitation cases; defined as knee osteoarthrosis present in IV degree, or functional class III onwards. This study evaluates cost and effectiveness of the use of Hylan G-F 20 vs. intraarticular steroids to withhold surgery in patients with severe knee osteoarthrosis. METHODS Cost-effectiveness analysis using a decision tree to simulate a hypothetical cohort behavior of patients with severe knee osteoarthrosis for a period of two years, from the perspective of the health service supplier. Costs were estimated using prices of 2008 and are expressed in US dollars (exchange rate of 11.14 pesos/ 1 US dollar). RESULTS With Hylan G-F 20, 94.6 % of patients did not require surgery during the analysis period vs. 50%, in the case of those under intraarticular steroid treatment. Expected treatment costs: Hylan G-F 20, $2081.0, and intraarticular steroids, $4593.2. The average cost-effectiveness of treatments: Hylan G-F 20, $2200.5 and intraarticular steroids, $9111.6. Incremental analysis shows Hylan G-F 20 as dominant alternative. Different sensitivity analyses corroborate the dominance relationship exercised by Hylan G-F 20 over the steroid treatment. CONCLUSIONS Hylan G-F 20 is a more effective and less expensive alternative than steroid treatment to withhold surgery in patients with severe knee osteoarthrosis.