Ca efflux from rat brain presynaptic nerve terminals (synaptosomes) was examined after loading the terminals with 45Ca during a brief depolarization, usually in media containing 20 microM Ca labeled with 45Ca, to assure a small (physiological) load. Efflux of 45Ca was very slow in the absence of external Na and Ca (approximately 0.5% of the load/s) and was greatly accelerated by Na and/or Ca (presumably Na+-Ca2+ and Ca2+-Ca2+ exchange, respectively). The dependence of 45Ca efflux on external Na was sigmoid, with a Hill coefficient of approximately 2.5; this implies that more than two external Na ions are required to activate the efflux of one Ca ion. The external Na (Nao)-dependent Ca efflux was inhibited by 1 mM external La, by low temperature (Q10 congruent to 2.3), and by raising external K (to depolarize the synaptosomes). With small Ca loads, the mitochondrial uncoupler, carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP), had negligible effect on either Ca uptake or efflux; with large loads (greater than or equal to 5 nmol/mg protein), however, FCCP reduced the depolarization-stimulated Ca uptake and increased the Nao-dependent Ca efflux. These effects may be attributed to reduction of mitochondrial Ca sequestration. Mitochondria do not appear to sequester much Ca when the loads are smaller (and more physiological). Estimations of Ca efflux indicate that approximately 20% of a small 45Ca load (approximately 0.75 nmol Ca/mg protein) may be extruded via Na+-Ca2+ exchange within 1 s; this corresponds to a net Ca efflux of approximately 110 pmol Ca X mg protein-1 X s-1.(ABSTRACT TRUNCATED AT 250 WORDS)