Band 3,. the major intrinsic protein of the human erythrocyte membrane, was specifically labeled with the covalent fluorescent probe eosin isothiocyanate. The lateral mobility of labeled bans 3 in the plane of the membrane under various conditions of ionic strength and temperature was examined by using the fluorescence photobleaching recovery technique. Low temperature (210C) and high ionic strength (46 mM NaPO4) favored immobilization of band 3 (10% mobile) as well as slow diffusion of the mobile fraction (diffusion coefficient D = 4 X 101 cm2 sec'1). Increasing temperature (370C) and decreasing ionic strength (13 mM NaPO4) led to an increase in the fraction of mobile band 3 (90% mobile) and a reversible increase in the diffusion rate of the mobile fraction (D = 200 X 10-11 cm2 sec'). The increase in the fraction of mobile band 3 was markedly dissociated, however, from the increase in the diffusion rate of the mobile fraction. Thus, the fraction of mobile band 3 always increased at higher ionic strength and lower temperature than the ionic strength and temperature at which the diffusion rate increased. This dissociation was manifested kinetically on prolonged incubation of ghosts at constant ionic strength and temperatures the diffusion rate of the mobile fraction increased slowly at first and much more rapidly after the initial lag period, whereas the fraction of mobile band 3 increased almost immediately to 90% and remained maximal for the duration of the experiment. Further, changes in diffusion rate with temperature were promptly and totally reversible, whereas increases in the mobile fraction were only slowly and partially reversible. These effects were shown not to be due to complete dissociation of spectrin, the major protein of the erythrocyte cytoskeleton, from the membrane. This evidence suggests control of band 3 lateral mobility by at least two separate processes. The process that determines the diffusion coefficient of the mobile band 3 is completely reversible, and it probably involves a metastable state of cytoskeleton structure intermediate between tight binding to the membrane and complete dissociation from it.Extensions of the fluid-mosaic model for membrane structure (1) have emphasized the importance of intramembranous (especially transmembranous) forces in the modulation of lateral mobility of integral membrane proteins (2, 3). Immunologic phenomena such as antibody-induced patching and capping and anchorage modulation (reviewed in ref.3) demonstrate a functional role for such forces in mammalian cell membranes. With reference to the human erythrocyte membrane, qualitative evidence for cytoskeletal control of integral membrane protein distribution and mobility has emerged from freeze-fracture electron microscopic studies (4-7). Direct examination of the lateral mobility of fluorescently labeled transmembrane proteins (chiefly band 3 and glycophorin) by the cell fusion (8) and photobleaching recovery (9) techniques has also suggested restricted mobility for these proteins.A definitiv...