The anion transport system of human red cells was isolated in vesicles containing the original membrane lipids and the 95 000 dalton polypeptides (band 3) by the method of Wolosin et al. (J. Biol. Chem. (1977) 252, 2419--2427). The vesicles have a functional anion transprot system since they display sulfate transport that is inhibited by the fluorescent probe 8-anilinonaphthalene 1-sulfonate (ANS) with similar potency as in red cells. The vesicles were labeled with the SH-specific probe fluorescein mercuric acetate (FMA). Labeling lowers FMA fluorescence, and is prevented or reversed by dithiothreitol, suggesting that the reaction is with a thiol group on the protein. Fluorescnece titrations show a maximum labeling stoichiometry of 1.3 +/- 0.4 mol FMA/mol 95 000 dalton polypeptide. The polarization of bound FMA fluorescence is high indicating that the probe is highly immobilized. Pretreatment with Cu2+ + o-phenanthroline under conditions that crosslink band 3 in ghosts decreases FMA labeling 50%. Differences in kinetics of FMA labeling in sealed and leaky vesicles suggest that the reactive SH group is located in the intravesicular portion of the protein (corresponding to the cytoplasmic surface of the red cell) and that FMA can cross the membrane. Inhibitors of anion transport have no effect on FMA labeling kinetics suggesting it is not transported via the anion transport system. Sulfate transport in the labeled vesicles remains fully functional. We detected self-energy transfer between bound FMA molecules by fluorescence depolarization. With excitation at 450--50 nm P decreases from 0.4, when less than half of the proteins are labeled, to 0.1 at saturation. This depolarization is not observed with red edge excitation (510--530 nm). Addition of 0.1% sodium dodecyl sulfate (SDS) changes P to 0.32, regardless of the excitation wavelength or degree of saturation with FMA. These results indicate that the band 3 proteins are close enough to allow energy transfer between fluorophores(Ro = 37.4 A), which does not occur upon red edge excitation or when the proteins are separated by SDS. We conclude that the functional anion transport system exists as a dimer or higher oligomer of band 3 proteins in these membranes, confirming previous suggestions derived using other methods. Future applications are discussed.