We reviewed the importance of measuring serum anticholinergic activity (SAA) in patients with Alzheimer's disease (AD). Since Tune and Coyle reported a simple method for assessing SAA using radioreceptor-binding assay, SAA is assumed to be the cumulative activity of parent medications and their metabolites and its relationship with delirium and cognitive functions has been debated. However, we evaluated the SAA in AD patients and SAA was correlated with prescription of antipsychotic medications, cognitive dysfunctions, severity of AD and psychotic symptoms, especially, with delusion and diurnal rhythm disturbance. From these results, we should not only pay attention to avoiding the prescription of medications with anticholinergic activity but also we speculated that AA appeared endogenously in AD and accelerated AD pathology. Moreover, there might be the possibility that SAA has predictive value for assessing the progressiveness of AD and as a biological marker for AD.