Controlled clinical studies have shown that local administration of morphine can significantly relieve acute postoperative pain. This analgesic effect is long-lasting (up to 48 h) and is mediated by peripheral opioid receptors. Experimental evidence shows that analgesic effects of peripheral opioids and the density of opioid receptors on peripheral sensory nerves increase with the duration of painful inflammatory processes. This study examines the analgesic effects of 1 mg of morphine injected into the arthritic knee joints of two groups of chronic pain patients (n = 23) suffering from osteoarthritis. Using a randomized, double-blind cross-over design, patients received either an intraarticular injection of morphine and intravenous saline (Group A, n = 13) or an intraarticular injection of saline and intravenous morphine (Group B, n = 10) during Phase I. Seven days later, patients crossed over to the opposite treatment (Phase II). During Phase I, intraarticular morphine resulted in significantly greater pain relief than intraarticular saline, and this effect was present at rest as well as during movement. The analgesic effect was surprisingly long-lasting and extended into Phase II, a carry-over effect that prevented the analysis of Phase II. No side effects were reported. The treatment of arthritic pain by peripherally acting opioids may be a promising alternative to currently available medications that have serious side effects.