Seventeen patients with painful osteoarthritis (OA) of the knee were given an intraarticular injection of corticosteroid while an equal number were given only the suspending vehicle. One wk later the corticosteroid group had significantly less pain than the placebo group but by 4 wk post-injection the difference between the 2 groups was significant. Post-injection flares occurred in both groups suggesting a steroid crystal-induced synovitis was not the cause. Because the additional pain relief afford by the steroid was temporary and the possible deleterious effects of intraarticular steroids, are still debated, their judicious use of OA is advised.