SUMMARY We used intracellular microelectrodes to study the effect of changes in extracellular calcium ion concentration [Ca 2+ ] o on the transmembrane potentials of canine cardiac Purkinje fibers in control Tyrode's solution and in the presence of agents thought to modify membrane permeability to potassium. In Tyrode's solution, decreasing [Ca 2+ ] o from 2.7 to 0.9 mM increased action potential duration measured at -60 mv (APD «,) and at full repolarization (APDioo) but did not significantly modify the normal linear relationship between cycle length and APD between cycle lengths of 500 to 4000 msec. We used 9-aminoacridine (9-AA) to decrease potassium permeability. At concentrations between 0.01 and 1.5 x 10" 5 M, 9-AA caused a concentration-dependent increase in APD-M and APD 1O o and a significant increase in the slope of the line relating APD to cycle length. (Schiitz, 1936;Cranefield and Hoffman, 1958;Surawicz et al., 1959) and that substitution of other divalent cations such as Sr 2+ for Ca 2+ delays repolarization (Brooks et al., 1955;Garb, 1951). After intracellular microelectrodes were introduced to the study of cardiac electrophysiology, it was shown in most studies that a decrease in [Ca 2+ ] o prolonged and an increase in [Ca 2+ ] o shortened the transmembrane action potential of both ventricular muscle fibers (Hoffman and Suckling, 1956) and Purkinje fibers (Hoffman and Suckling, 1956; Weidmann, 1955). Somewhat similar relationships between [Ca 2+ ] o and action potential duration were found for atrial muscle fibers (Hoffman and Suckling, 1956;Hoffman and Cranefield, 1960 field, 1960). Finally, a number of studies have shown for fibers from sheep and calf hearts that an increase in [Ca 2+ ] o shifts the plateau to more positive potentials and a decrease has the opposite effect (Reuter and Scholz, 1976;Beeler and Reuter, 1970).These effects of [Ca 2+ ] o on action potential duration and on the voltage level of the plateau were difficult to explain in terms of the finding for both Purkinje fibers and ventricular muscle that a significant inward current is carried by Ca 2+ through the secondary or slow inward channels (g si ) and that this current is responsible in large part for maintaining the membrane potential at depolarized levels after the initial action potential upstroke. In terms of these findings one would expect an increase in [Ca 2+ ] o to delay repolarization by increasing inward current and a decrease to have the opposite effect because of the change in driving force for Ca 2+ . An increase in action potential duration in the presence of high [Ca 2+ ] o has been seen only at quite low rates of stimulation (Niedergerke and Orkand, 1972); at usual rates of activation the opposite effect is observed. This apparently anomalous effect of Ca 2+ on action potential duration has been explained by assuming that changes in [Ca 2+ ] o result in changes in [Ca 2+ ]i and that the latter influences the conductance of the membrane by guest on