This, combined with the presence of early and delayed afterdepolarizations in trout cardiomyocytes suggests that the SR plays an important role in excitation-contraction coupling in the teleost heart and provides novel support for teleost cardiomyocytes as a relevant model for studies of the impact of genetic or pharmacological manipulation of SR function on cardiac arrhythmogenesis. The present study provides direct evidence for an active intervention of the SR in calcium handling on a beat-to-beat basis in trout. Thus, depletion of the SR calcium content strongly reduced the intracellular calcium transient, a value that is comparable to those reported in mammalian cardiomyocytes. However, the time course of inhibition of SR calcium uptake was slower than that observed in mammalian ventricular myocytes. This may at least partly be due to the 3–4 fold higher SR calcium load at steady state and a smaller fractional calcium release from the SR in trout atrial myocytes. As a result, it will take longer time from the onset of SERCA inhibition until the SR is depleted. In agreement with this it also takes a large number of stimulation pulses to fully reload the SR after calcium depletion with caffeine. The present results support previous indirect findings showing that the SR can contribute to the activation of contraction in trout ventricle. It also agrees with a large SR calcium CPI-613 storage capacity in trout cardiomyocytes, and SR calcium uptake-rates sufficiently high to allow SR calcium re-uptake on a beat-to-beat basis at physiological stimulation frequencies and temperatures. Furthermore, electrophysiological protocols designed to trigger and measure SR calcium release estimated it be near 50% of the total calcium transient. The results presented here suggest that the contribution could be even larger in trout atrial myocytes. The above findings strongly suggest that the SR does intervene in the excitation-contraction coupling in trout cardiomyocytes. In addition to this, the present study provides for the first time direct measurements of both spontaneous and triggered calcium release from the SR in trout and zebrafish, showing that the basic features of calcium sparks such as dimensions and duration are similar to those of calcium sparks recorded in mammalian and human cardiomyocytes under comparable conditions. Importantly, the frequency of calcium sparks and spontaneous calcium waves in trout atrial myocytes are also within the range of values previously reported in mammalian and human atrial myocytes under similar experimental conditions, supporting the notion that the key features of SR calcium release through the ryanodine receptor have been conserved phylogenetically. Interestingly, calcium sparks with similar properties were also observed in zebrafish ventricular myocytes. Since the L-type calcium current density is large in zebrafish, this finding shows that myocytes can have a high SR activity even though the ICa density is sufficiently large to provide for the calcium required for the activation of contraction. It is not clear why a previous study on isolated trout cardiomyocytes failed to detect calcium sparks under physiological conditions.