According to the occupation model, the observed sigmoidicity in receptor-binding studies frequently yields non-integer values for the number of molecules that bind per receptor, which may suggest inconsistencies of the model. Here, Bertil Tuk and Michael van Oostenbruggen re-examine the derivation of the model and pinpoint the origin of the observed inconsistencies, thereby demonstrating that these inconsistencies may lead to substantial error in estimates of receptor affinity, number of molecules that bind per receptor and cooperativity. A reformulated model allows more information to be derived from receptor-binding experiments, yielding integer estimates of the number of molecules that bind per receptor and quantitatively estimating the degree of cooperativity.