Isosmotic transport of fluid across epithelial cell layers occurs by intraepithelial mechanisms that are not fully understood. Newer methods of measuring water flows across epithelia with higher resolution should now permit some key issues regarding solute-linked water transport to be clarified. Unstirred-layer effects are not likely to be serious sources of error in these measurements with judicious choice of experimental conditions. Progress in ultrastructural stereology has shown that in the proximal tubule most of the transporting membrane is located in the basal aspects of cells, making models based on a hyperosmolar lateral compartment less relevant. The current models of simple transcellular osmosis, though appealing for this simplicity, fail to account for some major experimental findings. Experimental design and methodological limitations have not yet achieved rigorous testing of whether or not epithelia can produce a perfectly isosmotic absorbate without any transepithelial driving forces. A better understanding of the mechanism of translocation of water through the lipid bilayer, the plasma membrane proteins, and special membrane structures like the tight junctions would significantly contribute to our knowledge of the mechanisms and intraepithelial routes by which water is transported by epithelia.