The effects of equi-anaesthetic concentrations of halothane (HAL) and isoflurane (ISO) on myocardial performance, perfusion, oxygenation and lactate release were studied before, during and after a low-flow, global ischaemic insult in isolated, paced rat left heart preparations. An antegrade perfusion technique was used, where left atrial pressure (LAP) and mean aortic pressure (MAP) could be altered independently of each other. Aortic flow, coronary flow (CF) and PO2 in venous coronary effluent were continuously recorded and stroke volume, myocardial oxygen consumption (MVO2) and myocardial oxygen extraction as well as lactate release were calculated. The hearts were exposed for at least ten minutes to the perfusate without (control, n = 10) or with HAL (n = 10) or ISO (n = 10) at a MAP of 80 mmHg (10.4 kPa) and a LAP of 7.5 mmHg (1.0 kPa). After baseline measurements, MAP was reduced to 25 mmHg (3,2 kPa) for a total of nine minutes. Thereafter MAP was increased to 80 mmHg (10.4 kPa) for another nine minute period. During the whole experimental procedure, LAP was maintained at 7.5 mmHg (1.0 kPa) and heart rate at 325 beats per minute. In the pre-ischaemic control period, MVO2 was lower with HAL compared to ISO (P < 0.05) and control (P < 0.05). Stroke volume was also lower with HAL compared to control (P < 0.05). During hypoperfusion, lactate release was twice as high in the control group (P < 0.01) and with ISO (P < 0.01) compared to HAL.(ABSTRACT TRUNCATED AT 250 WORDS)