These studies test, using intravital microscopy (IVM), the hypotheses that perfusion effects on insulin-stimulated muscle glucose uptake (MGU) are ) capillary recruitment independent and ) mediated through the dispersion of glucose rather than insulin. For , capillary perfusion was visualized before and after intravenous insulin. No capillary recruitment was observed. For , mice were treated with vasoactive compounds (sodium nitroprusside, hyaluronidase, and lipopolysaccharide), and dispersion of fluorophores approximating insulin size (10-kDa dextran) and glucose (2-NBDG) was measured using IVM. Subsequently, insulin and 2[C]deoxyglucose were injected and muscle phospho-2[C]deoxyglucose (2[C]DG) accumulation was used as an index of MGU. Flow velocity and 2-NBDG dispersion, but not perfused surface area or 10-kDa dextran dispersion, predicted phospho-2[C]DG accumulation. For , microspheres of the same size and number as are used for contrast-enhanced ultrasound (CEU) studies of capillary recruitment were visualized using IVM. Due to their low concentration, microspheres were present in only a small fraction of blood-perfused capillaries. Microsphere-perfused blood volume correlated to flow velocity. These findings suggest that ) flow velocity rather than capillary recruitment controls microvascular contributions to MGU, ) glucose dispersion is more predictive of MGU than dispersion of insulin-sized molecules, and ) CEU measures regional flow velocity rather than capillary recruitment.