1. Whole-cell voltage-clamp measurements were made of the time- and voltage-dependent properties of the inwardly rectifying background potassium current IK1, in single myocytes from rabbit ventricle. The main goal of these experiments was to define the role of IK1 in the plateau and repolarization phases of the action potential (AP). 2. Action potentials from single ventricular myocytes were used as the command signals for voltage-clamp measurements. In these 'action potential voltage-clamp' experiments, IK1 was isolated from other membrane currents by taking the difference between control currents and currents in K(+)-free bathing solution. The results show that IK1 is small during the plateau, but then rapidly increases during repolarization and declines in early diastole. 3. Evidence of an important functional role for IK1 in AP repolarization was obtained by comparing the magnitude of IK1 and the rate of change of membrane potential (dVm/dt) in the same cell during the AP. The time courses of IK1 and dVm/dt during the AP were closely correlated, indicating that IK1 was the principal current responsible for final repolarization. 4. Rectangular voltage-clamp steps were used to study time- and voltage-dependent changes in IK1 at membrane potentials corresponding to the repolarization phase of the AP. 'Slow' relaxations or tail currents, lasting 100-300 ms, were consistently recorded when the cell was repolarized to potentials in the range -30 to -70 mV, following depolarizations between +10 and -10 mV. 5. The close correlation between the magnitude of the steady-state IK1 (in an external K+ concentration of 5.4 mM), which was outward for membrane potentials in the range -30 to -70 mV, and the magnitude of the tail currents, suggests that they resulted from a slow increase, or reactivation, of IK1. 6. The component of the slow tails due to reactivation of IK1 can be separated from a previously described component due to Na(+)-Ca2+ exchange since the IK1 component: (i) does not depend on the presence of the calcium current, ICa; (ii) can be recorded when internal EGTA (5 mM) suppresses large changes in [Ca2+]i; (iii) does not depend on the Na+ electrochemical gradient; (iv) is abolished in K(+)-free external solution; and (v) is not present in rabbit atrial myocytes, in which IK1 is very small. 7. The time- and voltage-dependent properties of IK1 revealed by these tail current experiments suggest that the measured magnitude of IK1 will be dependent on the voltage-clamp protocol.(ABSTRACT TRUNCATED AT 400 WORDS)