Ca2+-induced Ca2+ release (CICR) is fundamental to cardiac excitation-contraction (EC) coupling. In this process, ryanodine receptors (RyRs) in the sarcoplasmic reticulum (SR) are activated by a small Ca2+ influx via L-type Ca2+ channels (LCCs) in the surface membrane. It is thought that a group of RyRs within a ‘couplon’ (Franzini-Armstrong, et. al., 1999) are responsible for Ca2+ sparks, whose spatio-temporal summation produces the Ca2+ transient and contraction. It has been shown that Ca2+ transients occur with no detectable delay from the appearance of ‘trigger’ Ca2+ via LCCs and Ca2+ spark activation is relatively synchronous under physiological conditions (Cannell, et. al., 1994). Nevertheless, it is unclear how many LCCs contribute to the trigger Ca2+ during APs (Altamirano and Bers, 2007) and what process determines the variability in the timing of Ca2+ spark activation during EC coupling. Cardiac ventricular myocytes were obtained by enzymatic dissociation of Langendorff-perfused rat hearts. Animals were anaesthetised with a lethal dose of pentobarbital (140 mg/kg; I.P.) before any procedures were carried out (in accordance with the University of Auckland Animal Ethics Committee guidelines). Latencies of Ca2+ sparks and transients evoked by APs and voltage-clamp steps were measured. To determine the latency arising from LCC gating, the difference in Ca2+ spark timing between depolarising and repolarising steps was compared. During a repolarising step, Ca2+ influx is started by repolarisation from a very positive membrane potential (+ 100 mV) where LCCs are already open (e.g. Poláková, et. al., 2008). In the presence of 15 µM nifedipine, the mean Ca2+ spark latency was 5.6 ± 0.5 and 15 ± 2 ms (S.E.M., n = 11) for repolarising (no LCC delay) and depolarising steps to – 15 mV, respectively. The difference between these values (~ 8.8 ms) is similar to the expected latency for first LCC opening (Josephson, et. al., 2010). At + 5 mV, the difference in latencies decreased to ~ 4.5 ms, which suggests that CICR latency is dominated by the waiting time for LCC opening, rather than the increase in [Ca2+] produced by LCC openings (within this voltage range). Although it was only possible to examine a relatively narrow voltage range (due to the requirement to activate sufficient Ca2+ sparks from both depolarising and repolarising steps), the results were consistent with a simple model for CICR, where ~ 4 LTCCs were available in a couplon and each LCC could activate ~ 4 RyRs. The model can explain the observed latency of CICR during APs and suggests that CICR operates with a much higher coupling fidelity than would be expected from previous studies (e.g. Poláková, et. al., 2008).