The nature of the delayed blood-brain barrier (BBB) opening that occurs in rats subjected to forebrain ischemia by the technique of two-vessel (carotid) occlusion plus hypovolemic hypotension (2VO ischemia) was probed by examining the simultaneous, trans-barrier movement of two hydrophilic, normally poorly permeative solutes of markedly different molecular size: sucrose (MW = 342) and inulin (MW approximate to 5000). Pentobarbital-anesthetized, male, Sprague-Dawley rats (342-374 g) were subjected to 10 min of 2VO ischemia (tympanic temperature, 37.5-38.0 degrees C); 6 h later they were reanesthetized and, along with non-ischemic controls, injected i.v. with [C-14]sucrose and [H-3]inulin. Transfer constants (K(i)s) for blood-to-brain movement of the tracers and V(i)s (apparent initial volumes of tracer distribution) were determined for six brain regions by the multiple-time, graphical method(tracer circulation times from 3 to 30 min). V(i)s differed little or insignificantly between the two tracers, or between control and post-ischemic rats: the values did not suggest appreciable endothelial binding of either tracer that might lead to its uptake by adsorptive-phase endocytosis. In the controls, regional K(i)s +/-: S.E.M. (nl g(-1) s(-1)) for inulin ranged from 0.18 +/- 0.04 to 0.31 +/- 0.09 and were significantly lower(P < 0.01) than K(i)s for sucrose (1.53 +/- 0.16-1.91 +/- 0.29). The K-i ratio (sucrose/inulin) across brain regions (mean, 6.6; S.E.M., 0.6) was much lower than would be expected according to the concept that movement of most organic non-electrolytes across the intact BBB occurs by dissolution in and diffusion through endothelial cell plasma membranes, at a rate proportional to the lipid solubility and diffusivity of the solute. This finding is interpreted as indicating that a portion of the transfer of sucrose and inulin occurred by a mechanism other than dissolution-diffusion (e.g., via pores or vesicles). In the post-ischemic rats, K(i)s for bath tracers were elevated significantly (P < 0.01) in parietal cortex, striatum, hippocampus, and midbrain. The post-ischemic increases (Delta K(i)s) in these regions were greater for sucrose (1.90-3.31 nl g(-1) s(-1)) than for inulin (0.80-1.33). Across brain regions the ratio between sucrose Delta K-i and inulin Delta K-i averaged 2.9 (S.E.M., 0.2), a value significantly greater than the ratio of I that would be expected were the BBB opening due to an enhancement of micropinocytosis and vesicular transport. The correspondence of the mean Delta K-i ratio with the ratio of the free diffusion coefficients of the tracers (D-f,D-suc/D-f,D-inu = 2.9; water, 38 degrees C) suggests that the delayed opening of the BBB following 2VO ischemia involves the formation of trans- or paracellular, aqueous pores or channels.