Coding of binary mixtures by a population of olfactory receptor neurons in the spiny lobster (Panulirus argus) was examined. Extracellular single-unit responses of 50 neurons to seven compounds and their binary mixtures were recorded. The ability of a noncompetitive model with correction for binding inhibition to predict responses to mixtures based on responses to their components was compared with the predictive abilities of other models. This model assumes that different compounds activate different transduction processes in the same neuron leading to excitation or inhibition, and it includes a term quantifying the degree to which binding of an odorant to its receptor sites is inhibited by other compounds. The model accurately predicted the absolute response magnitude of the population of neurons for 13 of 15 mixtures assessed, which is superior to the predictive power of any of the other models. The model also accurately predicted the across neuron patterns generated by the binary mixtures, as evaluated by multidimensional scaling analysis. The results suggest that there is no emergence of unique qualities for binary mixtures relative to components of these mixtures.