Rapid activation of sarcolemmal L-type Ca channels (LCCs), Ca influx and activation of Ca-induced Ca release (CICR) from the sarcoplasmic reticulum (SR) are crucial to cardiac excitation-contraction (EC) coupling. The uniformity and duration of a Ca transient elicited by an action potential (AP) depends on LCC activation kinetics and ability of the consequent Ca influx to trigger Ca sparks. In contrast to some earlier work, recent studies have suggested that EC coupling ‘fidelity' is extremely low and that as many as 20-30 LCC openings are required to trigger a Ca spark. This result would suggest that the latency to Ca spark activation is dominated by LCC activation, (although this has not been shown). To investigate this idea, step voltage-clamp protocols were used to activate Ca transients and Ca sparks. In one series of experiments, LCCs were first activated by a pre-pulse. The latency of SR Ca release decreased with increasing depolarizing step potential, in qualitative agreement with the expected voltage-dependence of LCC activation. In contrast, the latency of SR Ca release increased with increasing step potential after a strong pre-pulse which should have removed the latency for LCC activation. The difference between these results reflects the influence of LCC gating kinetics on EC coupling latency. The results for Ca spark and Ca transient latency suggest that only a few LCC openings are required to trigger a Ca spark and that during a Ca transient, LCC availability was ∼3-fold larger. To test this interpretation, a computer model was constructed to simulate stochastic LCC openings, Ca appearance inside the dyad and Ca-dependent gating of ryanodine receptors. The model shows that the recorded Ca spark latencies can be reproduced when the number of LCCs that initiate SR Ca release is 1-2 and 5-6 for Ca transients.