Injectable ketorolac is an effective analgesic in ambulatory surgery patients. However, no studies have compared ketorolac with other NSAIDs in this setting. The analgesic efficacy of intramuscular ketorolac, rectal indomethacin and placebo was compared in healthy women undergoing gynaecological or breast surgery as outpatients. Ninety patients received 30 mg im ketorolac, 100 mg pr indomethacin or placebo in a prospective, randomized, double-blind manner. A standardized anaesthetic protocol was followed. Patients graded their pain on a 10 cm visual analogue scale in the recovery room, twice in the surgical day care unit and during the car ride home. The patients' postoperative fentanyl requirements, time to recovery milestones, and side effects were recorded. The placebo group received more fentanyl in the PACU but did not achieve the same pain relief as either of the NSAID-treated group (ketorolac 44 +/- 53 micrograms, indomethacin 39 +/- 55 micrograms, placebo 87 +/- 100 micrograms, P < 0.05). Patients who received an NSAID had less pain at 15 and 90 min (P < 0.05). The PACU stay was longer for the placebo group (ketorolac 50 +/- 13 min, indomethacin 49 +/- 12 min, placebo 62 +/- 35 min, P < 0.05). Time to ambulation was also longer in the placebo group (ketorolac 117 +/- 25 min, indomethacin 121 +/- 49 min, placebo 140 +/- 51 min, P < 0.05). However, no differences were observed between the two NSAIDS. Side effects were similar in all groups. We conclude that im ketorolac and pr indomethacin are equally effective analgesics in this group of patients.