For many years, the treatment of heart failure has focused, successfully, on the neurohumoral pathways, but recently more attention has again been given to the heart itself and ways to improve the phenotype of the failing cardiomyocyte. [Ca]i transients from myocytes of failing human hearts typically have a low amplitude and slow decline at normal frequencies.1–3 The slower decline has been attributed to a decreased Ca uptake into the sarcoplasmic reticulum (SR), as evidenced by decreased expression levels of the SR Ca-ATPase, SERCA, at both the mRNA and protein levels.4 Such deficiency of SERCA will lead to a decrease in SR content.5 Consequently, much attention has been dedicated to the potential treatment of heart failure by improving SERCA function either by pharmacological block of the inhibitory protein phospholamban (PLB)6 or by gene therapy targeted at SERCA itself or PLB. Such strategies have been successful in improving function in animal models7,8 and in isolated human myocytes.9 Although this seems a promising therapeutic venue, it should not mislead us into thinking that SERCA deficiency is the major (or even the only) defect responsible for the failing phenotype.8 In the last years, a number of other mechanisms have been identified that may contribute to the phenotype of human end-stage heart failure and may be targets for therapy. Upregulation of the Na-Ca exchange has been proposed as a compensatory mechanism for the decrease in SERCA function and could improve relaxation.10 However, upregulation of Na-Ca exchange may have negative consequences as well, such as prolongation of the action potential11 and further depletion of the SR. Experimental data on human myocytes suggest that the exchanger contributes to Ca loading during the latter part of the action potential.12 The exact function of the exchanger in heart failure is still unresolved, and whether any benefit of block or of further upregulation can be expected remains to be seen. Most recently, Marx et al13 reported that in end-stage human heart failure, the ryanodine receptor was hyperphosphorylated, which would lead to increased opening probability. As an isolated event, changes in ryanodine receptor opening probability would be expected to affect contraction only transiently,14 but in the setting of concomitantly decreased SERCA activity, loss of SR Ca is likely. With the limited availability of human tissue and the difficulty of obtaining proper controls, animal models have been most useful; similar decreases in SERCA activity and upregulation of Na-Ca exchange have been found in various animal models of heart failure (eg, Reference 15). Some of these studies have even led to novel concepts, as yet unexplored in human studies. In the failing rat heart, a decrease in SR Ca release was observed despite unchanged Ca current and SR Ca content.16 The authors speculated that this decrease in gain or efficiency of Ca release was related to local changes in the narrow cleft between sarcolemma and junctional SR resulting in a defective coupling between the ryanodine receptor and the Ca channel. In this issue of Circulation Research, another novel and exciting concept is advanced by Litwin et al.17 In a rabbit model of heart failure after myocardial infarction, they observe temporal and spatial heterogeneities in local Ca release events. Absent and delayed Ca sparks can account for not only the slower upstroke of the averaged whole-cell Ca transient but also for the slower relaxation, analogous to the late opening of single Ca channels contributing to the rate of inactivation of the whole-cell current. One of the perspectives offered by the authors is that we should revise our conventional approach, in which we consider mechanisms of systolic and diastolic dysfunction separately. Although any of the changes in human heart failure mentioned above are expected to affect both systolic and diastolic function, this is indeed not necessarily expected for isolated changes in Ca release. However, it is in line with recent clinical evidence indicating that, in heart failure, mostly both systolic and diastolic dysfunction are present.