The effect of cholinergic 'blockade' on human memory performance as a model for the effect of cholinergic depletion in clinical disorders was investigated. A wide range of memory functions was assessed in 70 subjects, using tests which were identical or closely similar to those which have previously been employed in clinical studies of Alzheimer and Korsakoff patients. In addition, a physiological measure of the degree of central cholinergic blockade was included, as well as measures of subjective arousal and objective attention. It was found that cholinergic blockade had no significant effect on the more passive aspects of primary (or 'working') memory, namely span tests and a measure of verbal short-term forgetting; in this, it contrasts with the marked deficits seen in Alzheimer-type dementia. On the other hand, cholinergic blockade produced impairment at a visuospatial short-term forgetting test, and at a verbal test in which the distractor task was made more difficult. On tests of secondary memory, cholinergic blockade produced a pattern similar to that seen in the anterograde amnesia of Alzheimer and Korsakoff patients, namely a pronounced impairment in learning verbal and visuospatial material, a 'normal' forgetting rate once learning had been accomplished, and relative preservation of the response to priming and of skill learning (procedural memory). Cholinergic blockade, however, did not produce a retrograde amnesia, nor did it affect the recall of temporal context or of long-established semantic knowledge. This pattern of results is compared with that obtained in previous studies of Alzheimer and Korsakoff patients.