Adenosine, an endogenous compound with a known antinociceptive effect when administered into the CNS, was applied in nine patients (21-65 years) by the peripheral intravenous route (70-130 micrograms.kg-1.min-1) as a replacement for peroperative opioids during inhalation anaesthesia for surgical procedures not requiring muscle relaxation. Lorazepam was given as premedication, thiopentone was used for induction, and succinylcholine facilitated intubation of the trachea. Anaesthesia was maintained with isoflurane [initial surgery endtidal concentration (ET) 0.88% (range 0.7-1%)] and nitrous oxide (60-70%) in oxygen. Adenosine infusion was initiated 5-10 min prior to surgery, and stopped close to or when isoflurane was terminated at the end of surgery. The duration of anaesthesia, adenosine infusion, and surgery were 120 min (range 80-165), 90 min (range 70-145), and 90 min (range 60-135), respectively. Spontaneous unassisted ventilation was maintained in all patients. Mean heart rate increased 10 beats.min-1 (range 0-35) upon induction of surgery, while systolic blood pressure was unaffected at 105 mmHg (range 85-120) (14 kPa (range 11.3-16.0)). Spo2 and ETCO2 were in the normal range. The isoflurane concentration was gradually reduced in most cases [mid-surgery ET 0.63% (range 0.5-0.8) and end-surgery ET 0.57% (range 0.3-0.8)]. Extubation and verbal communication were rapidly achieved after anaesthesia. The mean postoperative (24 h) opioid requirement was 4 mg (range 0-10 mg). These pilot cases suggest that systemic adenosine infusion may replace opioids during inhalation anaesthesia, and that adequate spontaneous ventilation can be achieved.