The accuracy of using 7- to 10-micron microspheres to measure blood flow to dog small intestinal villi was studied. These spheres appear to pass unimpeded through the afferent arterioles of the villus and lodge at the villus tip because 1) virtually all villus spheres were located at the tip, 2) flow determined by visually counting spheres did not differ significantly from flow determined from radioactivity of the sheared-off villus layer, 3) the size distribution of spheres in the villus and subvillus layers was virtually identical to that administered, indicating no premature impaction of the larger spheres, and 4) spheres lodging in the submucosa during vasoconstriction did not subsequently migrate to the villi during vasodilatation. Studies with 25-micron spheres indicated that 28% of the 7- to 10-micron spheres shunted through vessels greater than 10 micron. Double isotope studies showed that lodges spheres do not migrate and that the injection of 2--4 X 10(6) spheres does not alter villus flow. Thus, 7- to 10-micron microspheres pass to and lodge in villus vessels in proportion to capillary flow and should provide an accurate estimate of villus nutrient blood flow.