We have investigated the nature of Fura-2/AM loading into isolated perfused rat heart and the temporal and kinetic relationship between left ventricular [Ca2+]i dependent fluorescence and isovolumic pressure. The contribution of hydrolysed mitochondrial matrix Fura-2 fluorescence to that measured from the surface of the heart was estimated to be 43.9 +/- 5.5% by the addition of 100 microM Mn2+ to the perfusate. Maximum endothelial Fura-2 fluorescence ratio, estimated by the addition of 3 microM bradykinin to the perfusate, was found to constitute 33.6 +/- 2.7% of the maximum myocardial Fura-2 fluorescence ratio. Approximately 11.2% of the 340 nm surface fluorescence was insensitive to 20 mM Mn2+ in the presence of ionomycin (3 microM) and therefore indicates the degree of partial hydrolysis of Fura-2/AM. Thus, depending on the contribution of endothelial Fura-2 fluorescence at a physiological endothelial calcium concentration, cytosolic fluorescence may comprise between 11-45% of the total cellular fluorescence at 340 nm. Net tissue interference of the Fura-2 fluorescence ratio by NADH emission and myoglobin absorption remained unaltered, providing the oxygenation state of the tissue was unaltered throughout the experiment. The [Ca2+]i dependent fluorescence decay from peak systole was best fitted to a biexponential decay with fast and slow rate constants of 18.08 +/- 1.97 s-1 and 0.23 +/- 0.02 s-1, respectively. In addition, a phase shift was observed between temporal and kinetic measurements of the left ventricular isovolumic pressure and calcium dependent fluorescence traces during a contraction-relaxation cycle. We conclude that despite imperfect Fura-2/AM loading, the temporal and kinetic characteristics of intracellular [Ca2+] transients in normal isolated perfused rat heart are similar to those reported in more controlled preparations such as isolated myocytes and cardiac trabeculae.