The five muscarinic receptor subtypes (M1-M5) are characterized by seven helices that define a transmembrane cavity which serves as the binding pocket for agonists and antagonists. The five cavities appear to be topographically different enough to permit subtype selectivity among antagonists but not among classical agonists which tend to be smaller in size than antagonists. It was reasoned that synthesis of muscarinic agonists longer/larger than their classical counterparts might result in subtype selectivity. M1 subtype selectivity was found in a class of 1-azabicyclo[2.2.1]heptan-3-one, O-(3-aryl-2-propynyl) oximes. One of these, CI-1017, improved spatial memory of hippocampally deficient mice and nbM-lesioned rats at doses of 1.0-3.2 and 0.1-0.3 mg/kg, respectively, while producing parasympathetic side effects only at very high doses (100-178 mg/kg). Additionally, CI-1017 inhibited production of amyloidogenic A beta and increased secretion of soluble APP. Thus, CI-1017, besides treating AD symptomatically, may also retard its progression. CI-1017 has recently completed phase I clinical trials.