Exposure of the inner surface of intact red cells or red cell ghosts to Ca2+ evokes unitary currents that can be measured in cell-attached and cell-free membrane patches. The currents are preferentially carried by K+ (PK/PNa 17) and show rectification. Increasing the Ca2+ concentration from 0 to 5 microM increases the probability of the open state of the channels parallel to the change of K+ permeability as observed in suspensions of red cell ghosts. Prolonged incubation of red cell ghosts in the absence of external K+ prevents the Ca2+ from increasing K+ permeability. Similarly, the probability to find Ca2+-activated unitary currents in membrane patches is drastically reduced. These observations suggest that the Ca2+-induced changes of K+ permeability observed in red cell suspensions are causally related to the appearance of the unitary K+ currents. Attempts to determine the number of K+ channels per cell were made by comparing fluxes measured in suspensions of red cells with the unitary currents in membrane patches as determined under comparable ionic conditions. At 100 mM KCl in the external medium, where no net movements of K+ occur, the time course of equilibration of 86Rb+ does not follow a single exponential. This indicates a heterogeneity of the response to Ca2+ of the cells in the population. The data are compatible with the assumption that 25% of the cells respond with Pk = 33.2 X 10(-14)cm3/s and 75% with Pk = 3.1 X 10(-14)cm3/s. At 100 mM external K+ the zero current permeability of a single channel is 6.1 X 10(-14)cm3/s (corresponding to a conductance of 22 pS). Using appropriate values for the probability of a channel in the open state, we estimated that 25% of the cells in the population contain 11-55, and 75% of the cells 1-5 channels per cell that are activated in the time average (20 degrees C, pH 7.6).