Potter and Damiano recently assessed the hydrodynamic dimensions of the endothelial glycocalyx in vivo (mouse cremaster muscle venules) and in vitro (human umbilical vein and bovine aorta endothelium cultured in perfused microchannels) using fluorescent microparticle image velocimetry (Circ Res. 2008;102:770–776). Great discrepancy was observed, the glycocalyx presenting a zone of interaction extending 0.52 µm into the vessel lumen in vivo, but only 0.02 to 0.03 µm from cultured cells. In an accompanying editorial, Barakat cautioned that the difference in hydrodynamic interaction did not allow one to conclude that the cultured cells totally lack a physical cell surface layer capable of mechanotransduction (Circ Res. 2008;102:747–748). To stabilize the glycocalyx for electron microscopic investigation, we perfusion-fixed 6 human umbilical veins and confluent and nonconfluent cultures (5 each) of human umbilical vein endothelial cells (HUVECs) with lanthanum/glutaraldehyde solution. Ex vivo, the thickness of glycocalyx of umbilical vein endothelium averaged 878 nm. HUVECs in vitro presented a glycocalyx with a dense-zone thickness of only 29.4 nm, plus sparse filaments reaching out on average to 118 nm, there being no difference between the nonconfluent and confluent cells. Immunohistology demonstrated the presence of heparan sulfates and syndecan-1, main constituents of the glycocalyx, both ex vivo and in vitro. These results support the observed discrepancy between glycocalyx thickness in vivo and in vitro, now for one and the same type of human cell. The presence of heparan sulfates and syndecan-1 also on cultured cells may explain why mechanotransduction phenomena can be observed even with a nonmature glycocalyx.