The tissue uptake of extensively plasma-bound compounds is reportedly inconsistent with the conventional free-drug hypothesis limiting transport to unbound moiety in rapid intracapillary equilibrium with bound complex. Instead, protein-mediated/cell surface enhancement of dissociation has been postulated to occur in the microvasculature. This possibility was investigated by studying the passive transport of diazepam across the blood-brain barrier. Microdialysis probes placed within the vena cava and brain cortex were used to directly compare steady-state, interstitial unbound diazepam levels in both Wistar and genetically analbuminemic rats. The absence of albumin in the latter increased the unbound fraction of diazepam by almost fivefold; however, in both groups, the ratio of unbound concentrations in brain and blood at equilibrium was equal to unity. If enhanced dissociation occurred in the microvasculature, then the unbound brain level should have been greater than that in the systemic circulation. It is probable that earlier findings suggestive of protein-mediated transport reflect a nonequilibrium phenomenon. Comparison of the extent of diazepam's in vivo binding in blood by microdialysis to that estimated in vitro using conventional equilibrium dialysis with microcells showed good agreement, thus validating a widely accepted assumption of equivalency of these two values.