OBJECTIVES: We examined the efficacy, safety and patient satisfaction of intra-articular hyaluronic acid (HA) in patients with osteoarthritis of the knee. METHODS: One hundred patients with mild to moderate osteoarthritis of the knee entered a randomized blind-observer trial of 6 months HA vs placebo. Primary efficacy criteria were pain on walking, measured with a visual analogue scale, and the Lequesne Index. RESULTS: For pain on walking, a significant difference in favour of HA was found for completed patients at week 5, the end of the course of injections, and at month 6, the end of the study (P = 0.0087 and P = 0.0049, respectively). Further analysis using the Last Observation Carried Forward (LOCF) also showed a significant benefit favouring HA at month 6 (P = 0.0010). For the Lequesne Index, a significant difference in favour of HA was found at week 5 (P = 0.030) and at month 2 (P = 0.0431), but this was only of borderline significance at month 4 (P = 0.0528). Patients' global assessment of efficacy favoured HA at month 6 (P = 0.012). Improvement in other secondary criteria was generally superior in the HA group compared to placebo both at week 5 and month 6. Adverse events, mainly local injection site reactions, occurred in both groups with equal frequency. CONCLUSIONS: The study demonstrated that five weekly intra-articular injections of sodium hyaluronate (Hyalgan) were superior to placebo and well tolerated in patients with osteoarthritis of the knee with a symptomatic benefit which persisted for 6 months.