The rate of decline of free intracellular Ca concentration ([Ca]i) is a potentially useful index of the function of Ca transport systems. However, interpretations of these results may depend on multiple Ca transport systems and interaction with intracellular Ca binding sites. We measured [Ca]i in voltage-clamped ventricular myocytes isolated from rat hearts using indo 1 fluorescence. Conditions were chosen where [Ca]i decline was expected to depend almost exclusively on the sarcoplasmic reticulum Ca pump. The half time of [Ca]i decline (t1/2) decreased as the amplitude of the intracellular Ca (Cai) transient increased. This is not the result that would be expected from a transport system where the transport rate is a linear function of free [Ca]i. In this case the time constant of [Ca]i decline (tau) should be independent of the peak value of [Ca]i. This is also true if linear buffering of Cai is included. We develop a simple but more realistic theoretical framework where Ca transport rate and Ca binding both depend on free [Ca]i with Michaelis-Menten type functions. We demonstrate that the observed decline in apparent tau with increasing peak [Ca]i is entirely expected on theoretical grounds and over a wide range of characteristics for Ca transport and binding. We conclude that one cannot draw inferences about the intrinsic Ca transport function based on tau values unless the Cai transient has a comparable size.