The sleep changes induced in normal volunteers following the administration of scopolamine on 3 consecutive mornings resemble many of the abnormalities observed in the sleep of patients with primary depression: increased sleep latency and reduced rapid eye movement (REM) latency, total sleep time, and sleep efficiency. Furthermore, in a multivariate discriminant analysis--previously shown to distinguish the sleep records of depresed patients from those of normal controls and insomniac patients--the records from baseline nights were selected as normal and those after scopolamine as predominately depressed. Those observations suggest to us that muscarinic supersensitivity in normals may function as a pharmacological model for the sleep disturbances of depression.