The role of the cytosolic Ca (Ca i) transient in determining the configuration of the cardiac action potential was further investigated. For this purpose, rat ventricular myocytes loaded with the Ca 2+-sensitive fluorescent dye indo-1 were used. Myocytes were loaded with either 1) the acetoxy-methyl ester of indo-1 (indo-1 AM), which provides qualitative information about changes in the Ca (i) transient or 2) the Ca 2+- sensitive, free acid form of indo-1, which allows intracellular Ca (i) to be quantified. The magnitude of the Ca(i) transient was graded by various physiological and pharmacological interventions and membrane voltage or current was recorded using patch-type microlelectrodes. Earlier work in indo-1 AM cells had shown that phase-plane loops of membrane potential (Vm) vs indo-1 ratio from a stimulus train conformed to a common trajectory during the slow tail of repolarization, despite a beat-dependent decrease in the magnitude of the Ca(i) ransient. It was found that phase-plane loops of the Vm vs indo-1 ratio from spontaneous diastolic Cai oscillations also conformed to this trajectory. Experiments with indo-1 free acid, which was incorporated into the cell by inclusion in the microelectrode filling solution, showed that in rested beats peak Ca(i) an exceed 2 micrometer and then decrease to 0.4 - 0.6 micrometer in the steady state. Evidence was found that both electrogenic Na/Ca exchange and a Ca 2+-dependent ion channel current may participate in the modulation of the cardiac action potential by the Ca (i) transient.