An acetylcholine deficit remains the most consistent neurotransmitter abnormality found in Alzheimer's disease and various therapeutic agents have been targeted at this. In this study we investigated the action of Donepezil, a cholinesterase inhibitor that has few side-effects. In particular we set out to investigate whether muscarinic acetylcholine receptor (mAChR) availability influences the response to this therapy. We used the novel single-photon emission tomography (SPET) tracer (R, R)[(123)I]I-quinuclidinyl benzilate (R, R[(123)I]I-QNB), which has high affinity for the M1 subtype of mAChR. Regional cerebral perfusion was also assessed using technetium-99m hexamethylpropylene amine oxime. We investigated 20 patients on Donepezil treatment and ten age-matched controls. The results showed a reduction in (R, R)[(123)I]I-QNB binding in the caudal anterior cingulate in patients compared with controls and relatively high binding in the putamen and rostral anterior cingulate, suggesting a relative sparing of mAChR in these regions. The main finding of the study was that mAChR availability as assessed by (R, R)[(123)I]I-QNB binding did not distinguish responders from non-responders. Interestingly, we found that the extent of cognitive improvement showed no positive correlation with (R, R)[(123)I]I-QNB binding in any brain region but was inversely related to binding in the insular cortex. This suggests that, within the advised cognitive performance band for use of Donepezil, response is greater in those patients with evidence of a more marked cholinergic deficit. A larger study should investigate this.