Light-scattering measurements of osmotically induced changes in the size of rat liver microsomal vesicles pre-equilibrated in a low-osmolality buffer revealed the following. (1) The increase in extravesicular osmolality by addition of glucose 6-phosphate or mannose 6-phosphate (25 mM each) caused a rapid shrinking of microsomal vesicles. After shrinkage, a rapid swelling phase (t1/2 approx. 22 s) was present with glucose 6-phosphate but absent with mannose 6-phosphate, indicating that the former had entered microsomal vesicles, but the latter had not. (2) Almost identical results were obtained in the absence of any glucose 6-phosphate hydrolysis, i.e. with microsomes pre-treated with 100 microM-vanadate. (3) The anion-channel blocker 4,4'-di-isothiocyanostilbene-2,2'-disulphonic acid (DIDS) suppressed the glucose 6-phosphate-induced swelling phase. (4) The swelling phase was more prolonged as the glucose 6-phosphate concentration increased (t1/2 = 16 +/- 3, 22 +/- 3 and 35 +/- 4 s with 25 mM, 37.5 mM- and 50 mM-glucose 6-phosphate respectively). The behaviour of glucose-6-phosphatase activity of intact and disrupted microsomes measured in the presence of high concentrations (less than 30 mM) of substrate also indicated the saturation of the glucose 6-phosphate permeation system by extravesicular concentrations of glucose 6-phosphate higher than 20-30 mM. Additional experiments showed that vanadate-treated microsomes pre-equilibrated with 0.1 mM- and 1.0 mM-glucose 6-phosphate (and [1-14C]glucose 6-phosphate as a tracer) rapidly (t1/2 less than 20 s) released [1-14C]glucose 6-phosphate when diluted in a glucose 6-phosphate-free medium. The efflux of [1-14C]glucose 6-phosphate was largely prevented by DIDS, allowing an evaluation of the intravesicular space of glucose 6-phosphate of approx. 1.0 microliter/mg of microsomal protein.