Conventional semilogarithmic plots of radioisotope-retained [y(t)], by membrane vesicles during efflux, against time give a gross overestimate of the half-life (t 1/2) of the efflux, if a significant proportion of the radioisotope is retained at equilibrium. By using simulated efflux experiments, it is shown that apparent linearity of such plots evidenced by excellent linear correlation coefficients may not detect such error. 45Ca efflux from sarcolemma of rat myometrium further illustrates this point. Rat myometrium sarcolemmal vesicles were loaded at pH 6.27 and 7.07 with 1 mM CaCl2 and then diluted into solutions containing ethyleneglycol-bis(beta-aminoethylether)-N,N'-tetraacetic acid (EGTA) to result in 1 microM Ca2+ at the above pH values. Analysis using log y(t) vs. t plots gave t 1/2 values of 93.32 +/- 8.64 and 28.63 +/- 1.22 min at pH 7.07 and 6.27, respectively. However, log [y(t) - yeq] vs. t plots yielded t 1/2 values of 28.28 +/- 2.15 and 24.88 +/- 1.00 min. The discrepancy results from a larger equilibrium binding at the higher pH. Therefore it is recommended that the quantity of the isotope retained at equilibrium (yeq) be determined experimentally and then that the data be analyzed using log [y(t) - yeq] vs. t plots to obtain true estimates of t 1/2. This recommendation would be applicable to a wide range of studies.