The systemic administration of nonsteroidal anti-inflammatory agents has been shown to improve analgesia following arthroscopy of the knee joint. Ketorolac 60 mg, when given intra-articularly, provides better postoperative analgesia than an identical dose administered systemically. We compared the postoperative analgesic effect of ketorolac 10 mg given intravenously with 5 mg intra-articularly in 60 patients undergoing arthroscopy of the knee joint under general anaesthesia. Patients were randomly allocated in a double-blind manner to receive 0.25% bupivacaine 20 ml and ketorolac 5 mg intra-articularly (n = 27) or intravenous ketorolac 10 mg followed by 0.25% bupivicaine 20 ml (n = 30) at the end of surgery. There were no differences between the groups in terms of their physical characteristics or in the nature of procedure performed. There was no statistical difference between the two groups in time to first analgesia or postoperative visual analogue pain scores at 1, 2 and 4 h (p = 0.6). The median consumption of a standard analgesic was reduced in the intra-articular group in the second 24-h period but this did not achieve statistical significance (p = 0.08). Only five patients in total needed postoperative morphine. A reduced amount of locally applied ketorolac (5 mg) provides similar analgesia to a higher systemic dose (10 mg) following knee arthroscopy.