Intestinal microvascular permeability was studied in the isolated vascularly perfused small intestine of the rat by arterial injection of tracer molecules and collection of venous samples. The injection mixture contained a rhodamine-labeled dextran and a fluorescein-labeled dextran or free fluorescein. Pharmacokinetic analysis, based on statistical moment theory, of the tracer outflow concentration-time curve and the application of either the well-stirred model (WSM) or parallel tube model (PTM) was used to assess vasopermeability. The results indicate that the experimental system cannot be considered a pure WSM or a PTM. No different intrinsic clearance (Clint,i) values were found by applying the two models: Clint,i (in ml/min) = 1.23 +/- 0.14 (radius 0.5 nm); 0.44 +/- 0.09 (radius 1.4 nm); 0.31 +/- 0.08 (radius 2.2 nm); 0.02 +/- 0.01 (radius 6.0 nm); and 0 (radius 20.8 nm). Infusion of histamine (10(-5)-10(-3) M) and destruction of the endothelium via perfusion with distilled water increased the permeability for the tracers. We have established a technique for measurement of microvascular permeability characteristics in the rat small intestine. Histamine-induced changes and destruction of the endothelium can be detected in a quantitatively reliable way.