A key statement of the 1965 Monod-Wyman-Changeux (MWC) model for allosteric proteins concerns the distinction between the ligand-binding function (Y) and the relevant state function (R). Sequential models predict overlapping behavior of the two functions. In contrast, a straightforward experimental consequence of the MWC model is that for an oligomeric protein the parameters which characterize the two functions should differ significantly. Two situations, where R > Y and the system is hyper-responsive or where R < Y and the system is hypo-responsive, have been encountered. Indeed, the hyper-responsive pattern was first observed for the enzyme aspartate transcarbamoylase, by comparing Y with R monitored by a change in sedimentation. Extensions of the theory to ligand-gated channels led to the suggestion that, on the one hand, hyper-responsive properties also occur with high-affinity mutants. On the other hand, native channels of the acetylcholine neuronal alpha 7 receptor and low-affinity mutants of the glycine receptor can be interpreted in terms of the hypo-responsive pattern. For the ligand-gated channels, whereas R is detected directly by ion flux, ligand binding has rarely been measured and the formation of desensitized states may complicate the analysis. However, stochastic models incorporating both binding and channel opening for single molecules predict differences that should be measurable with new experimental approaches, particularly fluorescence correlation spectroscopy.