The L-type calcium current (ICa,L) plays an important role in excitation-contraction coupling of heart cells, as it forms the major trigger for Ca¥-induced Ca¥ release from the sarcoplasmic reticulum (SR) and provides Ca¥ for refilling of the SR (Callewaert, 1992;Barry & Bridge, 1993). The voltage relation of this current is bell shaped with an inward current maximum at about 10 mV and a reversal potential around 60 mV (McDonald et al. 1994). At the top of the action potential between 40 and 50 mV, when ICa,L is activated, ICa,L is already near its reversal potential and thus small. Moreover, the potential of maximum ICa,L is not reached before the repolarization of the cell. The obvious contradiction that ICa,L is important for excitationcontraction coupling on the one hand, but is small during the action potential on the other hand, leads to the idea of applying voltage-clamp pulses consisting of digitized action potentials to record the time course of ICa,L directly during the action potential. This so-called action potential clamp 1. During an action potential the L-type Ca¥ current (ICa,L) activates rapidly, then partially declines leading to a sustained inward current during the plateau phase. The reason for the sustained part of ICa,L has been investigated here. 2. In the present study the mechanisms controlling the ICa,L during an action potential were investigated quantitatively in isolated guinea-pig ventricular myocytes by whole-cell patch clamp. To measure the actual time courses of ICa,L and the corresponding L-type channel inactivation (fAP) during an action potential, action potential-clamp protocols combined with square pulses were applied. 3. Within the first 10 ms of the action potential the ICa,L rapidly inactivated by about 50%; during the plateau phase inactivation proceeded to 95%. Later, during repolarization, the L_type channels recovered up to 25%. 4. The voltage-dependent component of inactivation during an action potential was determined from measurements of L-type current carried by monovalent cations. This component of inactivation proceeded rather slowly and contributed only a little to fAP. ICa,L during an action potential is thus mainly controlled by Ca¥-dependent inactivation. 5. In order to investigate the source of the Ca¥ controlling fAP, internal Ca¥ homeostasis was manipulated by the use of Ca¥ buffers (EGTA, BAPTA), by blocking Na¤-Ca¥ exchange, or by blocking Ca¥ release from the sarcoplasmic reticulum (SR). Internal BAPTA markedly reduced the L-type channel inactivation during the entire action potential, whereas EGTA affected fAP only during the middle and late plateau phases. Inhibition of Na¤-Ca¥ exchange markedly increased inactivation of L-type channels. Although blocking SR Ca¥ release decreased the fura_2-measured cytoplasmic Ca¥ concentration ([Ca¥]é) transient by about 90%, it reduced L-type channel inactivation only during the initial 50 ms of the action potential. Thus, it is Ca¥ entering the cell through the L-type channels that controls the inactivation proc...