Intra-articular injection of long-acting insoluble corticosteroids produces rapid resolution of active arthritis in nearly all injected joints. Almost all of our information on the use of intra-articular corticosteroids in children comes from observational or retrospective analyses or, by inference, from studies in adult patients with arthritis. The duration of response has been found to vary according to the subtype of arthritis, the dose of injected steroids, the accuracy of injection, the duration of disease prior to injection, and possibly the age of the patient. Although the duration of follow-up in most studies has been short, intra-articular steroid therapy seems to be remarkably free of clinically important detrimental effects. Side effects are relatively uncommon and include subcutaneous atrophy and radiologically detectable structural changes or calcification. There is transient suppression of endogenous cortisol production, which may not be clinically important. Although intra-articular steroid therapy is most effective in pauciarticular juvenile rheumatoid arthritis, there are still no solid data to indicate whether it should be used earlier in the course of the disease instead of or along with systemic anti-inflammatory therapy. It has been suggested that repeated injection of the same joint decreases the likelihood of a favorable response. There are still many unanswered questions about how steroids exert their beneficial effects. Newer imaging techniques promise to provide insight into the mechanism of action and possibly to a more informed basis for the use of intra-articular steroids.