Viscosupplementation by hyaluronic acid (HA) injections is frequently used for local treatment of osteoarthritis (OA), due to ease of use and good tolerance. A profusion of linear or reticulated HA derivates are marketed, with varied characters and levels of evidence. Viscosupplementation has demonstrated moderate but significant efficacy (20%) versus placebo in terms of pain and function, with a high rate of responders (60-70%) in knee osteoarthritis. It allows reduced administration of opioid analgesics and NSAIDs, with improved risk/benefit ratio, and may delay joint replacement. Cartilage protection remains to be proven. Clinical efficacy shows 1-4 weeks' later onset than corticosteroids, but is maintained for 6 or even 12 months. Systematic association of corticosteroid and HA injection is not justified, and an interval has to be left before undertaking arthroplasty. Intra-articular injection of HA requires a skilled specialist, and may be difficult in a non-swollen joint; some tips and tricks may be helpful. In other joints than the knee, radiologic or ultrasound guidance is recommended. The efficacy of viscosupplementation is a matter of ongoing debate, after discordant findings in some meta-analyses. Some poor results may be due to inappropriate use of HA injections, poorly adapted to the patient's OA phenotype. Viscosupplementation is a treatment for chronic moderate symptomatic OA, and not for flares with joint swelling. Application in sport-related chondropathy has yet to be properly assessed. The optimal response profile remains to be determined. The ideal indication in the knee seems to be moderate femorotibial OA without swelling. Results have been generally disappointing in hip osteoarthritis but promising in OA of the ankle and shoulder (with and without rotator cuff tear). Further studies are needed to determine response profile and optimal treatment schedule, according to the joint.