Na+-dependent "binding" of acidic amino acids in brain plasma membranes was examined by procedures similar to those employed in earlier studies, using the metabolically inert D-[3H]aspartate as a probe. The "binding" showed characteristics similar to those described before in terms of affinity (KD, 400 nM), density of sites (Bmax, 300 pmol/mg protein), sensitivity to D,L-threo-3-hydroxyaspartate, and requirement for Na+. It turned out that the "binding" represents uptake into membrane-bounded saccules (which according to the inulin and H2O spaces constituted 3.4 microliters/mg protein and comprised about 50% of the volume of the sedimented membranes), rather than binding to the transport carrier. This conclusion is based on the observations that the "binding" of D-aspartate was released by osmotic shock; was abolished by thorough washing of membranes in H2O prior to assay, which removed endogenous contents of amino acids, and could be recovered by loading the washed membranes with glutamate; was reduced by prior freezing and thawing; was low on incubation at 0 degree C; had a bell-shaped time course similar to that reported for uptake; and had a slow rate of reversal compared to the apparent KD. True binding would have considerably lower apparent Bmax than the carrier-mediated uptake. This and its likely rapid rate of dissociation would make binding to the carrier difficult to detect by the methods used up to now.