1. The kinetic properties of the internal Na+ (Na+i)- dependent 45Ca2+ influx and external Na+ (Na+o)-dependent 45Ca2+ efflux were determined in isolated rat brain nerve terminals (synaptosomes) under conditions which the concentrations of internal Na+ ([Na+]i), external Na+ ([Na+]o), external Ca2+ (Ca2+]o), and external K+ ([K+]o) were varied. Both fluxes are manifestations of Na(+)-Ca2+ exchange. 2. Ca2+ uptake was augmented by raising [Na+]i and / or lowering [Na+]o. The increase in Ca2+ uptake induced by removing external Na+ was, in most instances, quantitatively equal to the Na+i-dependent Ca2+ uptake. 3. The Na+i-dependent Ca2+ uptake (measured at 1 s) was activated with an apparent half-maximal [Ca2+]o (KCa(o)) of about 0.23 mM. External Na+ inhibited the uptake in a non- competitive manner: increasing [Na+]o from 4.7 to 96 mM reduced the maximal Na+(i)-dependent Ca2+ uptake but did not affect KCa(o). 4. The inhibition of Ca2+ uptake by Na+o was proportional to ([Na+]o)2, and had a Hill coefficient (nH) of approximately 2.0. The mean apparent half-maximal [Na+]o for inhibition (KI(Na)) was about 60mM, and was independent of [Ca2+]o between 0.1 and 1.2mM; this, too, is indicative of non-competitive inhibition. 5. Low concentrations of alkali metal ions (M+) in the medium, including Na+, stimulated the Na+i-dependent uptake. The external Na+ and K+ concentrations required for apparent half-maximal activation (KM(Na) and KM(K), respectively) were 0.12 and 0.10mM. Thus, the relationship between Ca2+ uptake and [Na+]o was biphasic: uptake was stimulated by [Na+]o < or = 10 mM, and inhibited by higher [Na+]o. 6. The calculated maximal Na+i-dependent Ca2+ uptake (Jmax) was about 1530 pmol (mg protein) -1s-1 at 30 degrees C saturating [Ca2+]o and external M+ concentration ([M+]o), and with negligible inhibition by external Na+. 7. Internal Na+ activated the Ca2+ uptake with an apparent half-maximal concentration (KNa(i)) of about 20 mM and a Hill coefficient, nH, of approximately 3.0. 8. The Jmax for the Na+o-dependent efflux of Ca2+ from 45Ca(2+)-loaded synaptosomes treated with carbonyl cyanide p-trifluormethoxy-phenylhydrazone (FCCP) and caffeine (to release stored Ca2+ and raise the internal Ca2+ concentration ([Ca2+]i) was about 1800-2000 pmol (mg protein -1s-1 at 37 degrees C. 9. When the membrane potential (Vm) was reduced (depolarized) by increasing [K+]o, the Na+i-dependent Ca2+ influx increased, and the Na+o-dependent Ca2+ efflux declined. Both fluxes changed about 2-fold per 60 mV change in Vm. This voltage sensitivity corresponds to the movement of one elementary charge through about 60% of the membrane electric field. The symmetry suggests that the voltage-sensitive step is reversible. 10. The Jmax values for both Ca2P influx and efflux correspond to a Na+-Ca2+ exchange-mediated flux of about 425-575 jumol Ca2P (1 cell water)-' s-' or a turnover of about one quarter of the total synaptosome Ca2P in 1 s. We conclude that the Na+-Ca2P exchanger may contribute to Ca2P entry during nerve terminal depolarization; it is likely to be a major mechanism mediating Ca2P extrusion during subsequent repolarization and recovery.