40 Patients undergoing elective orthopaedic surgery of the lower limb randomly received either Buprenorphin sublingual (0,4 mg.) (Bsl) or Pentazocin i.m. (30 mg.) (Pim) for postoperative pain therapy. They were neither premedicated nor sedated. Analgesia was measured with a Visual Analogue Scale (VAS) and assessments by the anaesthetist and patient. Physiological measures were: Blood pressure, heart rate, arterial blood gases and serum cortisol. Side effects were registered. After Pim well known data were obtained: onset (10-30 min.) and duration (2 hours) of action and a significant pain reduction 20, 30 and 60 minutes after injection, whereas with Bsl onset was 60-120 min. and duration 6-7 hours of action and significant pain reduction was seen 120 and 180 min. after administration. 8 dropouts after Bsl up to 120 min. post applications are interpreted as a result of late onset or lack of analgesic action whereas the 13 drop outs 3 hours after Pim can be explained with the short lasting analgesic action of Pentazocin. Serum cortisol levels were better after Bsl, but did not correlate with the VAS. In Blood gas analyses, there were no group differences, especially no hypoxic or hypercapnic periods. There were no severe cardiocirculatory side effects, but significantly more hypertonic reactions and tachycardia after Pim. Other side effects were rare in both groups. The methods are discussed. Bsl has late onset but long duration of action with a higher efficacy than Pim. Bsl is strong and long acting. After an initial intravenous injection Bsl (0,4 mg.) should be administered at 6-hourly intervals.