The hypothesis of a Ca2+-induced Ca2+ release (CICR) from the sarcoplasmic reticulum (SR) is supported by experiments done in skinned cardiac cells (sarcolemma removed by microdissection). According to this hypothesis, the transsarcolemmal Ca2+ influx does not activate the myofilaments directly but through the induction of a Ca2+ release from the SR. The stimulus gating CICR is not a small change in free Ca2+ concentration (delta[free Ca2+]) outside the SR but a function of the rate of this change (delta[free Ca2+/delta t]). The initial relatively fast component of the transsarcolemmal Ca2+ current would trigger Ca2+ release; the subsequent slow component, perhaps corresponding to noninactivating Ca2+ channels, would load the SR with an amount of Ca2+ available for release during subsequent beats. Inactivation of CICR is caused by the large increase of [free Ca2+] outside the SR resulting from Ca2+ release, which inhibits further release. This negative feedback helps to explain that CICR is not all or none. During relaxation the Ca2+ reaccumulation in the SR is backed up by the Ca2+ efflux across the sarcolemma through Na+-Ca2+ exchange and the sarcolemmal Ca2+ pump. Computations of the Ca2+ buffering in the mammalian ventricular cell and of the systolic transsarcolemmal Ca2+ influx do not support the alternative hypothesis that this influx of Ca2+ is large enough to activate the myofilaments directly. Yet the hypothesis of a CICR can be challenged because of many problems and uncertainties related to the preparations and methods used for skinned cardiac cell experiments.