It is widely accepted that Ca transients in ventricular heart cells are produced by the summation of locally controlled release events. These release events are produced by couplons that reside in junctions. However we have recently shown that in isolated rabbit ventricular cells t-tubules are lost during the isolation procedure. This creates a significant number of RyR clusters that are non junctional and no longer exists as part of the couplon but are still able to release Ca (1). These non-junctional RyRs are far less numerous in tissue.. We studied mechanisms by which Ca is release is locally controlled from both couplons and non-junctional RyR clusters. Since t-tubules are known to be lost in various pathological conditions isolated rabbit cells provide a convenient model to study the nature of Ca transients in disease states. We have already established that, provided intracellular Na is present local release events are produced with a probability approaching 100%(2). We therefore investigated a possible mechanism that could explain this. With action potential shaped voltage clamp pulses we measured a significant Na current and an accompanying Ca transient using epiflourescence (3). However if the clamp pulse was preceded by a ramp, which inactivated all Na current the Ca transient, was reduced by 27%. We next investigated trans membrane Ca flux in isolated cells in which SR release flux was blocked. There was little difference in trans membrane Ca flux with and without the Na current. We therefore conclude that Ca current carries the majority of the flux. It therefore appears that a very small Ca flux mediated by NCX can have a large effect on SR Ca release if the SR is functioning. We further investigated this with a hypothesis first proposed by Maier et al (4). That Ca release can be mediated by a brain type (TTX sensitive Na channels). We measured SR release flux in the presence of 100 nM TTX and observed a 33 % reduction in release flux (similar to that produced by elimination of all Na currents). To explain this we propose that TTX sensitive Na currents are activated upon membrane depolarization and that that this primes the junctional cleft with Na. This Na in turn causes reverse NCX, which in turn primes the junction with Ca without triggering SR Ca release. Since we expect the relationship between SR Ca release and junctional Ca concentration to be sigmoid we imagine that priming the junction causes junctional Ca to increase along the foot of this sigmoid curve. Therefore when Ca currents are activated thy act on the steep portion of this curve which increases their coupling fidelity (5). Thus we view junctional priming as a mechanism that increases the coupling fidelity of Ca currents during an action potential. This ensures a high probability of activation of couplons and at least in part can account for the fact that local release events are observed to occur at high probability during action potentials in rabbits. We next investigated Ca transients in two dimensions using a Zeiss live 5 rapid scan head. We accomplished this by labeling cell membranes with di-8 anepps and using the Ca indicator fluo 4. Two-dimensional scans can be completed in 3.64 ms. Couplons were activated early and fairly synchronously. However Ca spread to non-junctional regions in a manner that was both temporarily and spatially inhomogeneous. In an attempt to understand whether the spread of Ca was simply due to diffusion from couplons or was due to activation of non-junctional release units we compared the spread of Ca in two dimensions with a model based on a reaction diffusion equation. The results suggested that non-junctional release units within 1 micron of the membrane (including t-tubules) were activated (presumably triggered by efflux from couplons). Beyond this distance we could not determine the extent to which non-junctional release unit were activated. These inhomogeneous Ca transient are expected to produce inhomogeneous sarcomere shortening with considerable weakening of contraction. Presumably this provides at least one explanation for the failure of contraction in end stage heart disease. We suggest therefore that local control of couplons does not simply occur by activation of Ca currents within the couplon but also involves Ca priming of the junctional cleft. Non-junctional RyRs are controlled by activation of local couplons but this control is more spatially extended. Moreover we suggest that a high probability of activation of couplons ensures optimal activation of non-junctional release units.