This study compared postoperative analgesic dispensation and measures relating to haemostasis following intravenous administration, in a randomized double-blinded manner, of either placebo or tenoxicam 20 mg to 30 women presenting for major gynaecological oncology surgery under a standardized, combined epidural/general anaesthetic technique. Pharmacokinetic disposition of tenoxicam in this patient cohort was also described. There was no objective or subjective alteration in haemostatic function or increase in blood loss, nor any deviation from the normal range of values. Postoperative analgesia during the first 48 hours was delivered to achieve a VAS endpoint of less than five on leg-raising, by a combination of a nurse-controlled low-dose background epidural infusion and patient-administered epidural bolus. Greater VAS variability was observed during the first four postoperative hours (P=0.08). The tenoxicam group self-administered significantly fewer epidural bolus doses to maintain satisfactory analgesia compared with the placebo group during the first 24 hours (P=0.004) and 48 hours (P=0.01) post-operatively. Similar differences between the groups in the total dose of the epidural bupivacaine/fentanyl mixture delivered were described (4h: P=0.148; 24h: P=0.033; 48h: P=0.001) (Figure 2). Despite surgery, transfusion and the use of a renal protective fluid administration strategy, tenoxicam disposition was not greatly different from that widely described for healthy volunteers. There were no significant side-effects and no adverse events attributable to tenoxicam. In this small study we have shown that tenoxicam administered preoperatively reduced the epidural analgesic requirements during the first 48 hours following major gynaecological surgery. There was no clinical or pathological evidence of haematological impairment following a single IV administration of tenoxicam 20 mg. Copyright © 2000 by Australian Society of Anaesthetists.