1. Activation of p38-mitogen-activated protein kinase (MAPK) has been implicated in the signalling cascade leading to protection by ischaemic preconditioning. This, however, is controversial and there is a plethora of conflicting data in the literature. Although many experimental differences may contribute to this, two in particular may be confounding: (i) the failure to account for p38-MAPK activation during aerobic perfusion; and (ii) the use of the anti-oxidant dimethylsulphoxide (DMSO) as the vehicle for the commonly used p38-MAPK inhibitor SB203580. We have investigated the effects of aerobic perfusion, ischaemia and preconditioning on p38-MAPK activation. In addition, we have used water-soluble SB203580 hydrochloride (SB203580.HCl) and DMSO to probe the role of p38-MAPK in preconditioning and ischaemic injury. 2. Activation of p38-MAPK in rat isolated hearts was assessed using a dual phosphospecific antibody during cannulation, aerobic perfusion and index, autolytic and preconditioned ischaemia. The effect of SB203580.HCl (10 mmol/L) in ischaemic preconditioning and ischaemia/reperfusion was tested using recovery of function and tetrazolium (TTC) staining as end-points. 3. Aerobic perfusion induced rapid activation (34% of maximal ischaemia-induced increase; P < 0.05) of p38-MAPK after 2 min that returned to baseline after 30 min. Index, autolytic and preconditioned ischaemia activated p38-MAPK, with index ischaemia peaking after 15 min (520% of basal; P < 0.05) before declining. SB203580.HCl blocked p38-MAPK activity, but did not block ischaemic preconditioning when bracketing the trigger phase and was not protective when given during ischaemia. 4. In the rat isolated heart, activation of p38-MAPK is neither a unique feature of preconditioning nor a prerequisite. Previous studies using SB203580 may have been complicated by failure to account for the activation of p38-MAPK by the protocol itself and the anti-oxidant properties of the most commonly used vehicle DMSO.