Résumé : Objectives: To compare the efficacy and safety of a "medium" molecular weight (MW) hyaluronan product (F60027, Structovial®) with a "high" MW (Hylan G-F20, Synvisc®). Methods: Prospective, randomised, multicentre, double-blind, active controlled, parallel-group study with a non-inferiority design. Patients with symptomatic KOA, global pain ≥40 mm (VAS, 0-100), Lequesne index (LFI, 0-24) score >7 and radiological Kellgren-Lawrence grade 2/3 were centrally randomised to receive F60027 or Hylan G-F20, administered via three weekly injections, with regular follow-up evaluations up to week 24 (W24). The primary outcome was LFI score change over 24 weeks. Secondary outcomes comprised pain VAS, quality of life, patient's and physician's global assessments, rescue medication consumption and OMERACT-OARSI responders rate. Results: 276 patients were analysed in the full analysis dataset (FAS), 236 in the Per Protocol dataset (PP). In the main efficacy analysis (PP), the difference of the LFl score change over 24 weeks between F60027 (-4.67 (0.27)) and Hylan G-F20 (-4.54 (0.28)) was 0.132 [95%CI: -0.598, 0.861 J which met the predefined non-inferiority margin. Analyses of secondary efficacy criteria showed clinically relevant improvements of all outcomes at W24 for each treatment on both PPIFAS populations. Changes of LFI score between baseline and W24 were -5.73 in the F60027 and -5.57 in the Hylan G-F20 group (PP dataset). Few local reactions were reported: 3.6% of patients in each group. Conclusion: F60027 and Hylan G-F20 were equally effective in reducing functional impairment and relieving pain in KOA patients, and well-tolerated. © Copyright Clinical and Experimental Rheumatology 2011.