The impermeant fluorescent probes (MIMAR reagents) described here permit the assessment of the lipid fluidity of individual membrane hemileaflets. They should also prove useful for examining the outer hemileaflets of the plasma membranes of intact cells. The observations, thus far, that normal human erythrocyte membranes have a characteristic asymmetry of fluidity, with the outer leaflet more fluid, correspond to prior findings with Mycoplasma, Newcastle Disease viral envelopes, and mouse LM cells. Hence, it is possible that the pattern is quite general in biological membranes. The particular lipid and protein components of the human-erythrocyte membrane that underly the fluidity asymmetry are unknown. The increased content of phosphatidylcholine in the outer leaflet and of the anionic phospholipids in the inner leaflet would be consonant with the fluidity difference. On the other hand, sphingomyelin, which tends to decrease fluidity, is localized mainly in the outer leaflet. Unknown at present is whether the cholesterol content of the two leaflets differs. From the results reported above, it is tempting to speculate that exogenously added cholesterol tends to localize in the outer leaflet, normally the more fluid leaflet, whereas endogenous cholesterol is more readily removed from the inner leaflet. This suggests, but clearly does not establish, that in the normal erythrocyte the cholesterol content of the inner leaflet exceeds that of the outer. Lastly, integral membrane proteins are expected to decrease lipid fluidity, and the usual pattern seen on freeze-fracture of large numbers of intra-membranous particles on the cytoplasmic face may signify a greater influence of protein in the inner leaflet. The hypothesis that perturbations of the fluidity of a given hemileaflet influence the membrane proteins (and their associated functions) in that leaflet is well-supported by the evidence described above. On the other hand, we understand less well the mechanisms by which lipid fluidity influences the proteins. For example, the decrease in sulfhydryl group reactivity of spectrin, actin, and Band 3 owing to cholesterol depletion (Table 7) may be due to a physical displacement of these proteins, as suggested by Borochov and Shinitzky. Why then does the reactivity of glyceraldehyde-phosphate dehydrogenase sulfhydryl groups increase under these conditions? There remains much to learn about membrane molecular mechanics and lipid-protein interactions. In such studies the impermeant MIMAR probes described here should prove useful.