1. The mechanisms that control release of Ca2+ from the sarcoplasmic reticulum (SR) of guinea‐pig ventricular cells were studied by observing intracellular calcium concentration ([Ca2+]i transients) and membrane currents in voltage‐clamped guinea‐pig ventricular myocytes perfused internally with Fura‐2. 2. Sarcolemmal Ca2+ current was identified through the use of tetrodotoxin (TTX) and Ca2+ channel antagonists (verapamil) and agonists (Bay K 8644). 3. Changes in [Ca2+]i attributable to release of Ca2+ from the SR were identified through the use of ryanodine, which abolishes the ability of the SR to release Ca2+. Ryanodine‐sensitive increases in [Ca2+]i could be elicited either by depolarization or by repolarization (from depolarizing pulses to relatively positive membrane potentials). 4. At appropriate voltages, it is the initial fast change in [Ca2+]i elicited by either depolarization or repolarization that is abolished by ryanodine, and is defined here as ryanodine sensitive. 5. The amplitude of the ryanodine‐sensitive [Ca2+]i transient elicited by depolarization had a bell‐shaped dependence on membrane potential with a maximum of about 500 nM at 10 mV, and with the upper minimum between 60 and 70 mV. Verapamil‐sensitive current activated over approximately the same potential range as the [Ca2+]i transient, with a peak amplitude at 10 mV, and a reversal potential of 65 mV. 6. When a holding potential of ‐68 mV and TTX (30 microM) were used, the most negative pulse potential at which activation of an inward current occurred was ‐49 mV while changes in [Ca2+]i occurred at ‐43 mV. 7. Ryanodine‐sensitive increases in [Ca2+]i elicited by repolarization (tail transients) were maximal for repolarization to 0 mV. Smaller changes in [Ca2+]i than maximal were elicited by repolarization to both more positive and more negative potentials than 0 mV. The peak amplitude of the verapamil‐sensitive tail currents elicited by repolarization increased continuously as the membrane was repolarized to potentials more negative than 60 mV. 8. Increasing depolarizing pulse duration beyond 10‐20 ms did not increase the amplitude of the [Ca2+]i transient, but prolonged it. 9. The experimental results are compared to the predictions of two theories on the mechanism of excitation‐contraction coupling: Ca2+‐induced release of Ca2+ (CICR), as it has been formulated from data in skinned cardiac cells, and a charge‐coupled release mechanism (CCRM), as it has been formulated to explain excitation‐contraction coupling in skeletal muscle. 10. Some of the results are clearly not consistent with certain features of a charge‐coupled release mechanism.(ABSTRACT TRUNCATED AT 400 WORDS)