1. Ischaemia was simulated in the isolated sheep cardiac Purkinje fibre and guinea-pig papillary muscle by immersing the preparations in paraffin oil. Ion-selective microelectrodes recorded potassium (Ks+) and pH (pHs) in the thin film of Tyrode solution trapped at the fibre surface while other microelectrodes recorded intracellular pH (pHi), membrane potential and action potentials (AP) (evoked by field stimulation), or membrane current (two-microelectrode voltage clamp in shortened Purkinje fibres). Twitch tension was also monitored. The paraffin oil model reproduced the salient characteristics of myocardial ischaemia, i.e. a decrease of twitch tension; a decrease of pHi and pHs; a rise in Ks+ (by 2-3 mM); a depolarization of diastolic membrane potential; considerable shortening of the AP (up to 30% within 4 min). 2. The sulphonylurea compounds, glibenclamide (200 microM) and tolbutamide (1 mM), known inhibitors of the KATP channel, completely blocked the ischaemic rise of Ks+ and prevented AP shortening. Ischaemic tension decline was notably less pronounced in the presence of sulphonylureas. 3. The ischaemic increase of slope conductance (Purkinje fibre) was prevented by 1 mM-tolbutamide and 200 microM-glibenclamide. 4. Sulphonylureas did not affect resting membrane potential, the AP or the current-voltage relationship under non-ischaemic conditions (this also indicates that ischaemic Ks+ accumulation is not fuelled by the background K+ current [iK1] which was shown, as expected, to be Ba2+ sensitive). 5. In a normally perfused preparation, reducing intracellular ATP by inhibiting glycolysis with 2-deoxyglucose (DOG) produced a similar AP shortening plus a membrane hyperpolarization, both of which were inhibited by tolbutamide or glibenclamide. The AP shortening was not related uniquely to the fall of pHi observed under these conditions since experimentally reducing pHi (by reducing pHo in the absence of DOG) lengthened rather than shortened the AP. 6. The possibility that the ischaemic rise in Ks+ might be the cause of AP shortening was excluded by the observation that, in a normally perfused Purkinje fibre, experimentally reducing pHi (by an amount similar to that seen during ischaemia) completely neutralized the AP-shortening effect of an elevated Ko+ (from 4.5 to 6.5 mM). Furthermore, the sulphonylurea-sensitive AP shortening seen during DOG treatment could not have been associated with a Ks+ rise since, in these particular experiments, the fibres were well perfused and diastolic membrane potential hyperpolarized.(ABSTRACT TRUNCATED AT 400 WORDS)