Although the gastric tension receptor has been characterized behaviorally and electrophysiologically quite well, its location and structure remains elusive. Therefore, the vagal afferents to the rat fundus (forestomach or nonglandular stomach) were anterogradely labeled in vivo with injections of the carbocyanine dye Dil into the nodose ganglia, and the nerves and ganglia of the enteric nervous system were labeled in toto with intraperitoneal Fluorogold injection. Dissected layers and cryostat cross sections of the fundic wall were mounted in glycerin and analyzed by means of conventional and laser scanning confocal microscopy. Particularly in the longitudinal, and to a lesser extent in the circular, smooth muscle layers, Dil-labeled fibers and terminals were abundant. These processes, which originated from fibers coursing through the myenteric ganglia and connectives, entered either muscle coat and then ran parallel to the respective muscle fibers, often for several millimeters. They ran in close association with the Fluorogold-labeled network of interstitial cells of Cajal, upon which they appeared to form multiple spiny appositions or varicosities. In the myenteric plexus, two different types of afferent vagal structures were observed. Up to 300 highly arborizing endings forming dense accumulations of small puncta similar to the esophageal intraganglionic laminar endings (Rodrigo et al., '75 Acta Anat. 92:79-100) were found in the fundic wall ipsilateral to the injected nodose ganglion. They often covered small clusters of myenteric neurons or even single isolated ganglion cells (mean = 5.8 neurons) and tended to extend throughout the neuropil of the ganglia. In a second pattern, fine varicose fibers with less profuse arborizations innervated mainly the central regions of myenteric ganglia. Camera lucida analyses established that single vagal afferent fibers had separate collaterals in both a smooth muscle layer and the myenteric ganglia. Finally, Dil-labeled afferent vagal fibers were also found in the submucosa and mucosa. Control experiments in rats with supranodose vagotomy as well as rats with Dil injections directly in the distal cervical vagus ruled out the possibility of colabeling of afferent fibers of passage. In triple labeling experiments, in conjunction with Dil labeling of afferents and Fluorogold labeling of enteric neurons, the carbocyanine dye DiA was injected into the dorsal motor nucleus of the vagus to anterogradely label the efferent vagal fibers and terminals. The different distributions and morphological characteristics of the vagal afferents and efferents could be simultaneously compared. In some instances the same myenteric ganglion was apparently innervated by an afferent laminar ending and an efferent terminal.(ABSTRACT TRUNCATED AT 400 WORDS)