Biological membranes define the structural and functional boundaries in living cells and their organelles. The integrity of the cell depends on its ability to separate inside from outside and yet at the same time allow massive transport of matter in and out the cell. Nature has elegantly met this challenge by developing membranes in the form of lipid bilayers in which specialized transport proteins are incorporated. This raises the question: is it possible to mimic biological membranes and create a membrane based sensor and/or separation device? In the development of a biomimetic sensor/separation technology, a unique class of membrane transport proteins is especially interesting-the major intrinsic proteins (MIPs). Generally, MIPs conduct water molecules and selected solutes in and out of the cell while preventing the passage of other solutes, a property critical for the conservation of the cells internal pH and salt concentration. Also known as water channels or aquaporins they are highly efficient membrane pore proteins some of which are capable of transporting water at very high rates up to 10(9) molecules per second. Some MIPs transport other small, uncharged solutes, such as glycerol and other permeants such as carbon dioxide, nitric oxide, ammonia, hydrogen peroxide and the metalloids antimonite, arsenite, silicic and boric acid depending on the effective restriction mechanism of the protein. The flux properties of MIPs thus lead to the question ifMIPs can be used in separation devices or as sensor devices based on, e.g., the selective permeation of metalloids. In principle a MIP based membrane sensor/separation device requires the supporting biomimetic matrix to be virtually impermeable to anything but water or the solute in question. In practice, however, a biomimetic support matrix will generally have finite permeabilities to both electrolytes and non-electrolytes. The feasibility of a biomimetic MIP device thus depends on the relative transport contribution from both protein and biomimetic support matrix. Also the biomimetic matrix must be encapsulated in order to protect it and make it sufficiently stable in a final application. Here, I specifically discuss the feasibility of developing osmotic biomimetic MIP membranes, but the technical issues are of general concern in the design ofbiomimetic membranes capable of supporting selective transmembrane fluxes.