S tandard and ion-sensitive microelectrodes were used to identify the basis of electrophysiologic changes that occur in canine cardiac Purkinje fibers superfused with "ischemic" solution (40 min) and then returned to standard Tyrode's solution. Maximum diastolic potential (EMDP) decreased (-92.6 +/- 2.4 to -86.0 +/- 4.0 mV; n = 19; P less than 0.001) during exposure to "ischemia," and after reperfusion, rapidly hyperpolarized to -90.0 +/- 4.7 (2 min) and then depolarized to -47.0 +/- 7.5 mV (10 min; P less than 0.001). No significant change in intracellular K activity (alpha ik) was noted throughout. Extracellular K activity (alpha ek) changed only during reperfusion, reaching a nadir at 5 min (3.5 +/- 0.4 to 2.6 +/- 0.5 mM, P less than 0.03), and thus can not account for the decrease in EMDP during reperfusion. Mean alpha iNa increased (8.7 +/- 1.3 to 10.9 +/- 1.9 mM; n = 10; P less than 0.01) during ischemia, but rapidly declined during reperfusion to 5.1 +/- 2.2 mM (10 min; P less than 0.01). Exposure to acetylstrophanthidin (4-5 x 10(-7) M) during the final 10 min of ischemia increased alpha iNa to 19.9 +/- 3.8 mM (n = 5), which was unchanged at 5 min of reperfusion. This suggests that Na-K pump inhibition during ischemia was minimal and that the pump was stimulated early during reperfusion, accounting for the initial transient hyperpolarization. Resting tension did not change significantly during exposure to ischemia; however, return to control Tyrode's solution caused a marked rise to 11.3 +/- 9.9 mg/mm2 (n = 13, P less than 0.001). This is consistent with a calcium overload state during reperfusion. The depolarization seen during reperfusion may result from activation of a Ca-activated, nonselective cation channel or enhanced electrogenic Na/Ca exchange.