We have used two bivalent ligands that bind IgE to study the relationship between the aggregation of receptors with high affinity for IgE (Fc epsilon RI) and the responses (receptor immobilization, Ca2+ influx, and degranulation) of rat basophilic leukemia (RBL-2H3) cells. One of these is a symmetric bivalent ligand, N,N'-bis[[epsilon-[(2,4-dinitrophenyl)amino]caproyl]-L-tyrosyl]-L- cystine ((DCT)2-cys), which binds specifically to the combining sites of a mAb anti-DNP IgE and efficiently cross-links cell surface IgE, but does not trigger significant degranulation or increases in intracellular Ca2+. Several lines of evidence, including lateral mobility measurements, indicate that this ligand preferentially forms stable cyclic complexes containing two (DCT)2-cys and two IgE. The second ligand is a mAb anti-IgE, B1E3, which causes lateral mobility changes consistent with dimerized IgE-Fc epsilon RI and also does not trigger increases in intracellular Ca2+ or degranulation. The two ligands together trigger robust responses. In the presence of B1E3, (DCT)2-cys causes immobilization of IgE-Fc epsilon RI in a broad concentration range; in a more narrow concentration range, it is a potent stimulant of changes in both degranulation and Ca2+. We have compared the dose-response curves for cellular activation to simulated IgE aggregation curves, i.e., curves that predict the equilibrium IgE aggregate size distribution as a function of the (DCT)2-cys concentration. Our results indicate that maximal cellular activation occurs at a much higher (DCT)2-cys concentration than maximal IgE aggregation. When IgE aggregation is maximal, almost all aggregated IgE is in cyclic dimers. Thus, cyclic dimers appear to be functionally ineffective, even after they have been cross-linked by B1E3. Aggregated IgE-Fc epsilon RI that is effective in stimulating a cellular response may have particular structural or dynamic properties that allow critical interactions for initiating the signaling cascade.