A specific action against myocardial reperfusion injury of the oxygen paradox type was recently characterized for halothane after anoxic perfusion in isolated rat hearts and isolated cardiomyocytes. In this study, we have characterized the protective effects of the clinically available inhalation anaesthetics during reperfusion after ischaemia. In isolated, isovolumically beating rat hearts perfused at a constant flow (10 ml min-1, PO2 80 kPa) and paced at 350 beat min-1, we determined left ventricular developed pressure (LVDP) and release of creatine kinase (CKR) as indices of myocardial performance and cellular injury, respectively. Seven control hearts underwent 30 min of no-flow ischaemia and 1 h of reperfusion. In the treatment groups, halothane, enflurane, isoflurane, sevoflurane or desflurane (each group n = 6) was added to the perfusion medium for the first 30 min of reperfusion at a concentration corresponding to 1.5 MAC in the rat. In the control group, cellular injury occurred at early reperfusion (peak CKR 283 (SEM 57) iu litre-1 at 10 min of reperfusion). Peak CKR to the coronary venous effluent was attenuated by all anaesthetics (halothane group 156 (45), enflurane group 134 (20), sevoflurane group 132 (20), desflurane group 159 (25) iu litre-1; each P < 0.05). Isoflurane did not differ from controls (303 (53) iu litre-1; P = 0.5). In the sevoflurane group, there was a delayed peak CKR after discontinuation of the anaesthetic at 30 min of reperfusion (260 (34) iu litre-1). Functional recovery was improved by all anaesthetics, but was seen much earlier with desflurane (LVDP 28 (3)% of baseline at 5 min reperfusion compared with halothane (6 (1)%), enflurane (11 (3)%), isoflurane (9 (6)%), sevoflurane (10 (2)%) and controls (3 (1)% of baseline)). At 30 min of reperfusion, recovery of LVDP was improved to a similar extent by all anaesthetics (halothane 30 (9)%, enflurane 36 (9)%, isoflurane 33 (5)%, sevoflurane 30 (5)%, desflurane 36 (4)% of baseline values) compared with controls (13 (5)%; each P < 0.05). All inhalation anaesthetics protected against myocardial reperfusion injury, but showed differences in attenuation of cellular injury and functional recovery. These differences may suggest different protective mechanisms.