1. Precise knowledge of the cytoplasmic Ca2+ buffering behaviour in intact human red cells is essential for the characterization of their [Ca2+]1-dependent functions. This was investigated by using a refined method and experimental protocols which allowed continuity in the estimates of [Ca2+]i, from nanomolar to millimolar concentrations, in the presence and absence of external Ca2+ chelators. 2. The study was carried out in human red cells whose plasma membrane Ca2+ pump was inhibited either by depleting the cells of ATP or by adding vanadate to the cell suspension.Cytoplasmic Ca2+ buffering was analysed from plots of total cell calcium content vs. (Lew, Tsien, Miner & Bookchin, 1982).This steady state is maintained with a minute total cell calcium content of 2-5 ,umol (1 cells)-', most of which could be extracted by external Ca2+ chelators from Ca2+-permeabilized cells (Harrison & Long, 1968;Bookchin & Lew, 1980;Engelmann & Duhm, 1987). Such small amounts of total and extractable Ca2P could be attributed to the scarcity of Ca2P-accumulating vesicles and to either the absence or unsaturation of cytoplasmic Ca2P buffers (Lew et al. 1982;Lew et al. 1985; Williamson, Puchulu, Penniston, Westerman & Schlegel, 1992 (Gardos, 1958;Tiffert, Spivak & Lew, 1988), binds to calmodulin (Scharif & Foder, 1978;Scharff, 1981), activates endogenous proteases and protease inhibitors (Glaser, Schwarz-Benmeir, Barnoy, Barak, Eshhar & Kosower, 1994;, promotes calpromotin binding to the cell membrane (Moore, Plishker & Shriver, 1990;Moore, Mankad, Shriver, Mankad & Plishker, 1991;Plishker, Chevalier, Seinsoth & Moore, 1992), and reacts with, or modifies the function of, a variety of other enzymes and cytoskeletal components (Lombardo & Low, 1994). Cytoplasmic Cah+ buffering in human red cells has been investigated in the past by a variety of methods, with similar overall results (Ferreira & Lew, 1976;Simons, 1982;Lew & Garcia-Sancho, 1989 In the analysis shown in Fig. 1 following usage, a will be treated as a dimensionless parameter. The true BAPTA concentrations were set at 5% (long-dash line) and 25% (short-dash line) below the nominal concentrations for both the sequential (Fig. 1A) and the simultaneous (Fig. 1B) protocols. In all cases, the mismatch generated false positive y-intercepts with apparent saturation patterns. In Fig. lA -s in human red cells 141BAPTA concentrations, respectively. In the simultaneous protocol (Fig. 1B) BAPTA was present, nominal BAPTA concentrations were used for the calculations. The conversion factor, f, used to minimize the difference between the plots with and without BAPTA, was then estimated as illustrated in Fig. 2A (A-D). BAPTA was not used in the last three experiments, in which the suspension media was pretreated with Chelex 100. The purity of BAPTA used for the remaining experiments in this set was 76%. The statistical parameters in each row correspond to the mean and standard errors derived from least mean squares fits to the two-buffer equation in each experiment (see Fig. 4). The...