To explain the equilibrium binding and binding kinetics of ligands to membrane receptors, a number of models have been proposed, none of which is able to adequately describe the experimental findings, in particular the apparent negative cooperativity of ligand binding. In this paper, a new model, the cluster-arranged cooperative model, is presented whose main characteristic is that it explains the existence of negative cooperativity in the binding of ligands to the receptor molecule. The model is based on our findings of agonist binding to A1 adenosine receptors and of ligand-induced clustering of these receptors on the cell surface. The model assumes the existence of two conformational forms of the receptor in an equilibrium which depends on the concentration of the ligand. In this way, negative cooperativity is explained by the transmission of the information between receptor molecules through the structure of the membrane. The model is able to predict the thermodynamic binding and binding kinetics of [3H]-(R)-(phenylisopropyl)adenosine to A1 adenosine receptors in the presence and absence of guanylyl imidodiphosphate. In the presence of the guanine nucleotide analogue, the linear Scatchard plots obtained for [3H]-(R)-(phenylisopropyl)adenosine binding are explained by the disappearance of cooperativity, thus suggesting that G proteins are important for the existence of negative cooperativity in ligand binding. Among other predictions, the model justifies early events in homologous desensitization since high ligand concentrations would lead to the saturation of the receptor in a low-affinity conformation that does not signal. Our model can likely explain the behavior of a number of heptaspanning and tyrosine-kinase receptors exhibiting complex binding kinetics.