Capillary permeability-surface area products for 86Rb, [51Cr]EDTA (mol. wt. 357), [57Co]cyanocobalamin (mol. wt. 1353) and [125I]insulin (approximate mol. wt. 6000) have been measured using the single-circulation, multiple-tracer dilution technique in the in situ perfused submandibular salivary gland during parasympathetic nerve stimulation, close-arterial bradykinin infusion and following chronic duct ligation. 2. In glands with a natural blood supply, permeability-surface area for 86Rb and [51Cr]EDTA increased during parasympathetic stimulation, but this was shown to be related to the concomitant increase in blood flow rather than to a change in capillary permeability or in surface area. 3. In glands perfused at constant flow, parasympathetic stimulation led to a decrease in permeability-surface area for EDTA (-19.1 +/- 5.2%, mean +/- S.E., n = 5, P less than 0.05) cyanocobalamin (-12.3 +/- 6.0, n = 12, P less than 0.05), and insulin (-15.3 +/- 4.8, n = 11, P less than 0.02). It is suggested that this may be the result of a redistribution of flow from the acinar microcirculation to a less permeable ductal vasculature. 4. Bradykinin infusion had no significant effect on permeability-surface area for EDTA and cyanocobalamin in perfused glands. 5. In perfused glands, ligation of the submandibular duct for 3--12 days reduced permeability-surface area (ml.min-1.g-1) for [51Cr]EDTA from 5.26 +/- 0.60 (mean +/- S.E., n = 9) to 4.20 +/- 0.12 (n = 4, P less than 0.30), [57Co]cyanocobalamin from 3.22 +/- 0.12 (n = 48) to 2.02 +/- 0.08 (n = 15, P less than 0.001) and [125I]insulin from 1.52 +/- 0.07 (n = 39) to 0.72 +/- 0.23 (n = 11, P less than 0.001).