The microvascular architecture of rat duodenal mucosa was investigated by scanning electron microscopy of vascular corrosion casts, and of fractured tissue specimens in which the interstitial matrix had been digested by acid/enzyme treatment. Light microscopy of cleared tissues, in which the vasculature had been injected with silicone elastomer, was used to define the spatial organization of the larger supply/drainage vessels. Within the mass of Brunner's glands, possibly because of its compact nature, it proved impossible to define differing microvascular networks related to particular functional regions of the glands (acini and ducts). The microvascular bed of Brunner's glands is, however, clearly a separate vascular bed from that ofthe rest of the duodenal mucosa, and is arranged essentially in parallel to it; it is argued that this separation of venous outflow from the glands may be important in maintaining the HCO3" protection of the duodenal surface from acid erosion. Microvascular architectural differences in the villi and intestinal gland regions, compared to more distal rat small intestine, include an apparently denser villus subepithelial capillary network origin of villus venules closer to the villus tip, and frequent direct drainage of the pericryptal network into adjacent venules.As part of a continuing study relating the microvascular organization to differing regional functions of the gut wall, we have investigated the microvascular architecture of the proximal duodenum in the rat, with particular attention to that of the duodenal (Brunner's) glands. As a result of recent studies of gastric microvascular architecture, we have recently proposed an intramucosal vascular transfer of HCO3_ in oxyntic mucosa This paper is dedicated to the memory of Ms Jan Nowell, who died recently. Ms Nowell, formerly of the Electron Microscope Laboratory, University of California, Santa Cruz, was a pioneer in the application of the corrosion vascular casting-SEM technique, and a personal friend of B.G. Her contributions in teaching and research will be sorely missed. which might assist in protection of the gastric surface from back-diffusing H+ ions (6, 7). In the duodenum, secretion of HCO3" by Brunner's glands must be associated with acidification of the adjacent interstitial fluid: one might predict that venous blood effluent from this region would not pass towards the surface epithelium of the duodenum (in contrast to the equivalent situation in the stomach), since acidified blood is unlikely to do other than compromise duodenal mucosal protective mechanisms against luminal acid.Thus the current study paid particular attention to the venous drainage pattern of Brunner's glands, and the possible blood supply routes to the duodenal surface, in order to test whether such predictions, based on functional considerations, might prove correct.