IT has been proposed that the regulation of the ionised calcium concentration inside many excitable and non-excitable cells is mediated by a Na–Ca countertransport mechanism where the Na electrochemical gradient provides the energy to maintain low levels of intracellular Ca2+ (refs 1, 2). Two recent observations, however, provide evidence consistent with the idea that Ca extrusion in squid axons is mediated, at least in part, by an ATP-driven Ca pump of the type observed in red blood cells3,4. First, in the absence of a Na gradient across the membrane there is an ATP-dependent ‘uphill’ extrusion of Ca which persists in the absence of external Na, Ca and Mg (ref. 3). And second, in Ca-injected unpoisoned axons, 50–90% of the Ca efflux can occur as an ‘uncoupled’ flux—that is, Ca efflux not accompanied by the uptake of Na, Ca or Mg (ref. 5). The existence of two apparently separate mechanisms for Ca transport, Na–Ca exchange and an ATP driven Ca-pump, raises the question of whether, in normal conditions, both mechanisms participate in the maintenance of the physiological pCa. To be of any use in controlling the pCa of the nerve, the mechanism responsible must be able to operate at normal Ca2+ concentrations, <10−7 M (ref. 6). We now present evidence showing (1) that in squid axons with physiological levels of Ca2+i most of the Ca efflux can be accounted for by an ATP-dependent system with high affinity for Ca2+i and ATP and which can operate in the complete absence of external Na and Ca; and (2) that a Ca transport mechanism with a low affinity for Ca2+i and ATP and extremely sensitive to Nai, Nao and Cao can contribute substantially to the total Ca efflux only at Ca2+i well above the physiological Ca2+ concentrations.