Aims The prevalence of positive carotid sinus massage (CSM) in asymptomatic elderly persons and related risk of syncope are poorly known. Methods and results We examined the ability of CSM to predict the risk of profound bradycardia or syncope in 30 patients with Alzheimer's disease (AD) and no history of syncope, before treatment with cholinesterase inhibitors. Carotid sinus massage was repeated at 1, 2 and 8 months after onset of cholinergic therapy. The tong-term incidence of syncope was observed over a 20-month follow-up. Among the study patients (mean age = 80 +/- 6 years, 83% women), 10 had positive CSM, with prolonged ventricular standstill in two of them (6.7%). The response to CSM was not predictive of an increased risk in bradycardia-mediated syncope, though syncope occurred in three patients during long-term follow-up, related to orthostatic hypotension in one and to undetermined causes in the others. Cholinergic therapy in all patients, or drugs that slow cardiac conduction, administered to 40% of the responders to CSM and 45% of the non-responders, or the presence of a bundle branch block did not affect the response to CSM or incidence of syncope. Conclusions A positive CSM was (i) observed in 30% of elderly persons with AD and no history of syncope, and (ii) not predictive of an increased risk of bradycardia-mediated syncope.