Nicotinic acetylcholine receptors are made up of homologous subunits, which are encoded by a large multigene family. The wide number of receptor oligomers generated display variable pharmacological properties. One of the main questions underlying research in molecular pharmacology resides in the actual role of this diversity. It is generally assumed that the observed differences between the pharmacology of homologous receptors, for instance, the EC 50 for the endogenous agonist, or the kinetics of desensitization, bear some kind of physiologic relevance in vivo. Here we develop the quite challenging point of view that, at least within a given subfamily of nicotinic receptor subunits, the pharmacologic variability observed in vitro would not be directly relevant to the function of receptor proteins in vivo. In vivo responses are not expected to be sensitive to mild differences in affinities, and several examples of functional replacement of one subunit by another have been unravelled by knockout animals. The diversity of subunits might have been conserved through evolution primarily to account for the topologic diversity of subunit distribution patterns, at the cellular and subcellular levels. A quantitative variation of pharmacological properties would be tolerated within a physiologic envelope, as a consequence of a near-neutral genetic drift. Such a "gratuitous" pharmacologic diversity is nevertheless of practical interest for the design of drugs, which would specifically tackle particular receptor oligomers with a defined subunit composition among the multiple nicotinic receptors present in the organism.