The local calcium release flux signals (calcium spikes) evoked by membrane depolarization were recorded at high temporal resolution (2000 lines s−1) in isolated ventricular myocytes of male rats, using combination of scanning confocal microscopy and the patch-clamp technique. The kinetic properties of calcium spikes were investigated. The time course of calcium spike activation could be described reliably by a model with higher-order (n= 3) kinetics, but not by a first-order exponential process. A model of calcium spike with calcium release termination coupled to its activation was preferential to a model with the release termination independent of its activation. Three fluorescent calcium dyes (OG-5N, fluo-3, and fluo-4) were compared for calcium spike measurements. Experimental measurements as well as simulations showed that the occurrence and latency of calcium spikes could be measured faithfully with all indicators, while the kinetics of calcium spikes was reliably traced only with OG-5N. Calcium spikes evoked by a step depolarization from −50 to 0 mV commenced with a mean latency of 4.1 ± 0.2 ms and peaked 6.7 ± 0.2 ms later. Their full amplitudes were normally distributed. The activation time constant of calcium spikes was 3.1 ± 0.1 ms, and the time constant of termination was 5.5 ± 0.2 ms. A negative correlation was observed between the observed amplitude of calcium spikes and their time constant of activation, but there was no correlation between their observed amplitude and time constant of termination, in agreement with the concept of steep calcium-dependent activation and fateful inactivation of calcium release flux.