All our advanced, severe cases of Alzheimer's disease have dramatic cholinergic cell losses in the nucleus basalis of Meynert even after correction for cell or nucleoli shrinkage. There is a good correlation between choline acetyltransferase activity and "healthy" cell number in the nucleus basalis of Meynert. Half of the Alzheimer disease cases have markedly reduced cortical choline acetyltransferase activity in spite of preserved nucleus basalis of Meynert choline acetyltransferase activity, suggesting a deficiency of cortical origin and/or of axonal transport in Alzheimer disease. The relationship between cell loss in the various sub-divisions of the nucleus basalis of Meynert and plaque counts in corresponding and non-corresponding projection areas of the cortex has also been examined. Globally, this relation appears more obvious when cell loss in a sub-division of the nucleus basalis of Meynert is compared to plaque counts in its cortical projection area. However, the relation is discontinuous with few or no data to document the intermediary stages of the process, probably reflecting the severity of our Alzheimer disease cases.