OBJECTIVE A brief period of ischaemia (5 min) and reperfusion (5 min), prior to a longer period of ischaemia and reperfusion, has been shown to reduce the extent of injury (necrosis, arrhythmias, or postischaemic contractile malfunction) caused by a subsequent longer period of ischaemia and reperfusion. Adenosine has been identified as a factor in the protection afforded against regional tissue necrosis by such preconditioning. The aim of this study was to assess the role of adenosine in preconditioning induced protection of postischaemic function in the globally ischaemic isolated rat heart. METHODS The ability of global ischaemia to precondition against postischaemic contractile malfunction was first confirmed in the isolated ejecting rat heart preparation. Hearts (n = 6 per group) were perfused aerobically (37 degrees C, paced at 350 beats.min-1) for 20 min, at the end of which contractile function was measured. This was followed by 10 min of Langendorff perfusion (control group) or 5 min of global ischaemia plus 5 min of Langendorff reperfusion (preconditioned group). The hearts were then subjected to 20 min of global ischaemia (37 degrees C) and 35 min of reperfusion (15 min Langendorff and 20 min ejecting); function was then reassessed. RESULTS Postischaemic recovery of aortic flow was 26(SEM 8)% in the control group v 57(4)% in the preconditioned group (p < 0.05). To assess whether exogenous adenosine could mimic this protection, the experiments were repeated with the 5 min period of ischaemic preconditioning replaced by 5 min of aerobic Langendorff perfusion with adenosine-containing buffer (100, 50, or 10 mumol.litre-1). No protection of postischaemic function was observed in any of the adenosine treated groups. In further experiments, we assessed whether ischaemic preconditioning persisted in the presence of the A1/A2 adenosine antagonist, 8 (p-sulphophenyl) theophylline (8-SPT). Since pacing was not used in these studies, the ability of ischaemia to precondition the myocardium was again confirmed; the protocol was then repeated with 8-SPT (10 mumol.litre-1) present in the perfusate throughout. Although 8-SPT depressed recovery in both control and preconditioned hearts it failed to abolish the protective effects of ischaemic preconditioning. CONCLUSIONS There is no evidence from these results to support the involvement of adenosine to any major extent in preconditioning induced protection of postischaemic contractile function in the isolated rat heart.