In studies that attempt to measure intracellular calcium [( Ca2+]i) in the intact heart with the calcium indicator indo 1-AM, a fundamental assumption is that the signals report changes in myocyte [Ca2+]i. We studied isolated perfused rabbit hearts loaded with the calcium probe indo 1-AM and recorded surface fluorescence of the left ventricle during continuous excitation at 360 nm. In cells containing indo 1, an increase in [Ca2+]i is associated with an increase in fluorescence intensity at 400 nm, a decrease in intensity at 500 nm, and an increase in the 400:500 nm ratio. Beat-to-beat fluorescence transients were recorded from the surface of the heart coincident with contraction, indicating that a component of the fluorescence signals is derived from beating myocytes. To evaluate the potential contribution of endothelial cells, we compared the response to increases in [Ca2+]o or bradykinin (10(-5) M). In response to an increase of the [Ca2+] in the perfusate from 0.6 to 3.0 mM, left ventricular developed pressure and +dP/dt increased with a simultaneous increase in the [Ca2+]i-sensitive 400:500 nm ratio. Perfusion with the endothelial cell agonist bradykinin caused no change in left ventricular isovolumic peak systolic pressure or left ventricular dP/dt, whereas bradykinin evoked an immediate elevation in both the diastolic and systolic levels of the [Ca2+]i-sensitive 400:500 nm ratio. In additional experiments with indo 1-loaded isolated beating myocytes, superfusion with bradykinin had no effect on either the fluorescence [Ca2+]i transients or contractility. In contrast, superfusion of indo 1-loaded cultured endothelial cells with bradykinin caused the elevation of [Ca2+]i within seconds. Fluorescence microscopy of unstained frozen tissue sections from indo 1-loaded hearts also suggested the presence of more intense microvascular endothelial cell indo 1 fluorescence relative to that observed in myocytes. These experiments provide evidence that a component of [Ca2+]i-sensitive fluorescence of whole hearts loaded with indo 1 is contributed by nonmyocyte sources, including endothelial cells. These results also raise the caution that the abrupt rise of [Ca2+]i that has been observed during the initial phase of ischemia in whole hearts loaded with indo 1 may be partly derived from endothelial cells rather than myocytes.