This report introduces a novel method to load indo-1 "free acid" selectively into the cytosol of cardiac myocytes, presumably by diffusion through momentarily permeable gap junction sites during mechanical dissociation after low-Ca2+ collagenase treatment. Calibration of indo-1 fluorescence in these cells has been accomplished after subtracting average autofluorescence (AF) from time-matched non-indo-loaded cells, taking into account apparent changes in cell AF due to indo-1. There is wide variation in the degree of uncertainty of individual intracellular Ca2+ concentration ([Ca2+]i) determinations among cells, related principally to differences in cellular indo-1 content, to nonlinear aspects of the [Ca2+]-to-fluorescence ratio relationship, and to the uncertainty in the AF subtraction. Consequently, a quantitative estimate of uncertainty also may be employed in formulating weighted estimates of cytosolic [Ca2+]i. The following [Ca2+]i values in rat ventricular cells (nM; in 1 mM bathing extracellular Ca2+ concentration, 25 degrees C) are given as weighted means +/- 95% confidence intervals (unweighted values in parentheses): 138 +/- 5 (136 +/- 6, n = 44) in quiescent cells, 435 +/- 74 (482 +/- 76, n = 43) at the [Ca2+]i-transient peak during 0.5 Hz steady-state stimulation, and 760 +/- 124 (1,027 +/- 250, n = 42) at the [Ca2+]i-transient peak, postrest. Moreover, these peak [Ca2+]i values fall near the steepest portion of the force-Ca2+ curve (from intact cardiac muscle), consistent with sensitive inotropic regulation and maximal contractile reserve.