Muscarinic receptor blockade in humans induces a transient memory deficit claimed to mimic aspects of Alzheimer's disease (AD). AD is also strongly associated with a specific blood flow abnormality in parietotemporal cortex; we previously showed that, despite induction of a dementia-like state, scopolamine does not produce these blood flow changes. In the present study, we administered both muscarinic and nicotinic receptor blockade (using scopolamine and mecamylamine) to seven elderly healthy subjects and measured subsequent changes in cognition and cortical perfusion, using the 133Xe inhalation method to quantify regional cerebral blood flow (rCBF). Results confirmed earlier findings of scopolamine-induced memory deficit and frontal cortex flow reduction. Only mecamylamine, however, produced a perfusion deficit in parietotemporal cortex. All effects were transient and dose-dependent. These findings demonstrate the safety and feasibility of differential and combined blockade of nicotinic and muscarinic cholinergic blockade in human subjects. Furthermore, the nicotinic antagonist mecamylamine yields rCBF changes similar to those seen in AD, despite producing only minimal cognitive effects on its own. The rCBF and behavioral manifestations in AD may therefore reflect the functional loss of nicotinic receptors in addition to alterations in other receptor systems.