1. Guinea-pig ventricular myocytes, isolated enzymatically without the aid of special media, show a similar sensitivity to the calcium paradox as Langendorff-perfused hearts. 2. Measurement of the intracellular activities of Na+ and Ca2+ ions, with a suction-type ion-sensitive microelectrode at rest, during calcium depletion and during inhibition of the Na+ pump (under both current and voltage clamp) yield values similar to those obtained from multicellular preparations and from isolated myocytes by other means. 3. In voltage-clamped myocytes bathed by media free of divalent cations, an inward sodium current that flows through the L-type Ca2+ channels, the rate of rise of aiNa and the strength of the contraction induced by return to normal Tyrode solution, show a similar bell-shaped dependence on the membrane potential during the period of Ca2+ deprivation. 4. The rise in aiNa that occurs in Ca(2+)-free, Mg(2+)-free media, induces an outward current which is composed of currents due to activation of the Na+ pump and K+ channels. 5. On Ca2+ repletion the loading of the cells with Ca2+ does not generate an inward current and the contracture can be reduced, in a dose-dependent way, by the introduction of BAPTA into the sarcoplasm from the solution in the voltage electrode. When [Ca2+]i is buffered by added BAPTA, the estimated amount of Ca2+ which can enter on Ca2+ repletion is sufficient to bind up to 10 mM of the BAPTA. This change in concentration is similar to that expected from the rise and fall in aiNa, seen on Ca2+ depletion and repletion, if a 3 Na+:1 Ca2+ exchange is responsible for the Ca2+ influx. 6. These data offer support for the so-called intracellular sodium hypothesis for the origin of the calcium paradox in the heart. As the effects of Ca2+ repletion can be prevented by clamping the membrane potential so that aiNa does not rise, the contribution of the other effects of Ca2+ depletion to the initiation of the calcium paradox would seem to be less important.