Defects in excitation-contraction coupling have been reported in failing hearts, but little is known about the relationship between these defects and the development of heart failure (HF). We compared the early changes in intracellular Ca(2+) cycling to those that underlie overt pump dysfunction and arrhythmogenesis found later in HF. Laser-scanning confocal microscopy was used to measure Ca(2+) transients in myocytes of intact hearts in Wistar-Kyoto (WKY) rats and spontaneously hypertensive rats (SHRs) at different ages. Early compensatory mechanisms include a positive inotropic effect in SHRs at 7.5-9 mo compared with 6 mo. Ca(2+) transient duration increased at 9 mo in SHRs, indicating changes in Ca(2+) reuptake during decompensation. Cell-to-cell variability in Ca(2+) transient duration increased at 7.5 mo, decreased at 9 mo, and increased again at 22 mo (overt HF), indicating extensive intercellular variability in Ca(2+) transient kinetics during disease progression. Vulnerability to intercellular concordant Ca(2+) alternans increased at 9-22 mo in SHRs and was mirrored by a slowing in Ca(2+) transient restitution, suggesting that repolarization alternans and the resulting repolarization gradients might promote reentrant arrhythmias early in disease development. Intercellular discordant and subcellular Ca(2+) alternans increased as early as 7.5 mo in SHRs and may also promote arrhythmias during the compensated phase. The incidence of spontaneous and triggered Ca(2+) waves was increased in SHRs at all ages, suggesting a higher likelihood of triggered arrhythmias in SHRs compared with WKY rats well before HF develops. Thus serious and progressive defects in Ca(2+) cycling develop in SHRs long before symptoms of HF occur. Defective Ca(2+) cycling develops early and affects a small number of myocytes, and this number grows with age and causes the transition from asymptomatic to overt HF. These defects may also underlie the progressive susceptibility to Ca(2+) alternans and Ca(2+) wave activity, thus increasing the propensity for arrhythmogenesis in HF.