1. Slow inward tail currents attributable to electrogenic sodium-calcium exchange can be recorded by imposing hyperpolarizing voltage clamp pulses during the normal action potential of isolated guinea-pig ventricular cells. The hyperpolarizations return the membrane to the resting potential (between -65 and -88 m V) allowing an inward current to be recorded. This current usually has peak amplitude when repolarization is imposed during the first 50 ms after the action potential upstroke, but becomes negligible once the final phase of repolarization is reached. The envelope of peak current tail amplitudes strongly resembles that of the intracellular calcium transient recorded in other studies. 2. Repetitive stimulation producing normal action potentials at a frequency of 2 Hz progressively augments the tail current recorded immediately after the stimulus train. Conversely, if each action potential is prematurely terminated at 0.1 Hz, repetitive stimulation produces a tail current much smaller than the control value. The control amplitude of inward current is only maintained if interrupted action potentials are separated by at least one full 'repriming' action potential. These effects mimic those on cell contraction (Arlock & Wohlfart, 1986) and suggest that progressive changes in tail current are controlled by variations in the amplitude and time course of the intracellular calcium transient. 3. When intracellular calcium is buffered sufficiently to abolish contraction, the tail current is abolished. Substitution of calcium with strontium greatly reduces the tail current. 4. The inward tail current can also be recorded at more positive membrane potentials using standard voltage clamp pulse protocols. In this way it was found that temperature has a large effect on the tail current, which can change from net inward at 22 degrees C to net outward at 37 degrees C. The largest inward currents are usually recorded at about 30 degrees C. It is shown that this effect is attributable predominantly to the temperature sensitivity of activation of the delayed potassium current, iK, whose decay can then mask the slow tail current at high temperatures. 5. Studies of the relationship between the tail current and the membrane calcium current, iCa, have been performed using a method of drug application which is capable of perturbing iCa in a very rapid and highly reversible manner. Partial block of iCa with cadmium does not initially alter the size of the associated inward current tail. When iCa is increased by applying isoprenaline, the percentage augmentation of the associated tail current is much greater but occurs more slowly.(ABSTRACT TRUNCATED AT 400 WORDS)