This article describes two prospective, randomized, double-blind clinical trials designed to investigate this. Trial 1 compared a conventional local anaesthetic agent (100 mg bupivacaine) injected intra-articularly (i.a.) with a control (normal saline) and 1 mg of i.a. morphine. No significant difference was noted in the first 4 hours between the groups with respect to visual analogue pain (VAS) scores. However, at 6 and 24 hours, the group of patients who received 1 mg i.a. morphine recorded lower pain scores and required less supplementary analgesia. Trial 2 assessed the dose response relationship for i.a. morphine comparing 5 mg intravenous (i.v.) morphine (control) with 1 mg and 5 mg i.a. morphine. At early time points (1, 2, and 4 hours) similar VAS pain scores were recorded for both 5 mg i.v. morphine and 5 mg i.a. morphine, both significantly lower than the group receiving 1 mg i.a. morphine. At 6 and 24 hours, 5 mg of i.a. morphine produced significantly lower pain scores, less analgesic requirement, and less sleep disturbance on the first postoperative night than the other groups. It can be concluded from these two studies that 5 mg i.a. was the most effective analgesic following knee arthroscopy.