Red blood cells (RBCs) from patients with sickle cell anemia and thalassemia carry abnormal accumulations of molecular Fe(III) at the cytosol/membrane interface. The avidity of the red cell membrane for this iron has not been defined. Using open ghost membranes prepared from sickle RBC, we examined the ability of membrane-associated Fe(III) to resist removal by 15 chelators representing a 40-log range of affinities for Fe(III). Efficacy of chelators was compared with literature values for their idealized affinity for iron as represented by the cummulative stability constant (β(n)), their effective stability constant reflecting affinity under biologic conditions (K(eff)), and an indicator of their ability to chelate Fe(III) in the presence of an insoluble phase of iron (K(sol)). Deferoxamine, a very high affinity chelator having log β(n) = 30.6, was found to be the lowest affinity chelator able to remove RBC membrane Fe(III). Regardless of chelator β(n), only those agents able to preserve log K(eff) ≤12 were able to do so, indicating that the membrane's effective avidity for Fe(III) is on the order of 1012. Additional confirmation that membrane avidity for Fe(III) is extremely high is found in the observation that only chelators having log K(sol) >0 were effective. Potential physiologic iron chelators in cytoplasm of pathologic red cells are unable to prevent or reverse iron accumulation on the membrane because they do not have sufficiently high affinity for iron. These data argue that RBC membrane Fe(III) is truly pathologic.