Most of the calcium that activates cardiac contraction is released from the sarcoplasmic reticulum (SR). This release is triggered by the entry of calcium into the cell via the L-type Ca channel leading to opening of the ryanodine receptor (RyR). This process is known as Ca induced Ca release (CICR). It has been suggested that, in addition to CICR there may also be a voltage dependent mechanism of Ca release (Ferrier & Howlett, 2001). However other work fails to find such a mechanism (Trafford & Eisner, 2003) and has suggested that it may result from incomplete inhibition of calcium entry (Piacentino III et al. 2000).The amount of Ca released depends on the Ca content of the SR; an increase of SR content increases the amount released. Stable contraction of the heart therefore depends on control of SR content (Eisner et al. 2000). We will review evidence showing that control of SR content depends on the effects of Ca release on surface membrane Ca fluxes. Specifically an increase of SR content increases SR Ca release and this (1) increases Ca efflux from the cell on Na-Ca exchange and (2) decreases Ca entry into the cell via the L-type Ca current. Both of these effects will decrease cell and thence SR Ca content (Trafford et al. 1997). This simple homeostatic mechanism has many consequences. For example it predicts that manoeuvres which only increase the open probability of the RyR will have no maintained effect on the amplitude of the systolic Ca transient. Experimentally we find that such an increase of open probability results in a transient increase of the amplitude of the Ca transient accompanied by a decrease of SR Ca content (Trafford et al. 2000). Correspondingly, a decrease of open probability produces a transient decrease of the Ca transient amplitude and an increase of SR content (Overend et al. 1998). Under some clinical conditions the cardiac output shows ‘alternans’. At a cellular level this is seen as alternate large and small systolic Ca transients (D’az et al. 2002). We are investigating the hypothesis that this alternation is due to breakdown of the homeostatic mechanism described above. We find that, under alternating conditions there is an increase in the steepness of the relationship between SR Ca content and Ca efflux from the cell. The factors responsible for this will be described.