Calcium-45 efflux was measured in squid axons whose internal solute concentration was controlled by internal dialysis. Most of the Ca efflux requires either external Na (Na-Ca exchange) or external Ca plus in alkali metal ion (Ca-Ca exchange; cf. Blaustein & Russell, 1975). Both Na-Ca and Ca-Ca exchange are apparently mediated by a single mechanism because both are inhibited by Sr and Mn, and because addition of Na to an external medium optimal for Ca-Ca exchange inhibits Ca efflux. The transport involves simultaneous (as opposed to sequential) ion counterflow because the fractional saturation by internal Ca (Cai) does not affect the external Na (Nao) activation kinetics; also, Nao promotes Ca efflux whether or not an alkali metal ion is present inside, whereas Ca-Ca exchange requires alkali metal ions both internally and externally (i.e., internal and external sites must be appropriately loaded simultaneously). ATP increases the affinity of the transport mechanism for both Cai and Nao, but it does not affect the maximal transport rate at saturating [Ca2+]i and [Na+]o; this suggest that ATP may be acting as a catalyst of modulator, and not as an energy source. Hill plots of the Nao activation data yield slopes congruent to 3 for both ATP-depleted and ATP-fueled axons, compatible with a 3 Na+-for-1 Ca2+ exchange. With this stoichiometry, the Na electrochemical gradient alone could provide sufficient energy to maintain ionized [Ca2+]i in the physiological range (about 10(-7) M).