Positron emission tomography (PET) with 18F-fluorodeoxyglucose (FDG-PET) thus far rarely has been used to advance the development of new treatments for Alzheimer's disease (AD). Now that FDG-PET with standard acquisition protocols for dementia is widely available, change in cerebral glucose metabolism is a feasible outcome variable for clinical drug trials. Individual analysis of FDG-PET results also might prove valuable. FDG-PET can detect metabolic changes very early in the course of AD and identify subjects for earlier treatment. FDG-PET reliably distinguishes AD from frontotemporal dementia so that only those most likely to benefit are enrolled in trials. Finally, objectively identifying phenotypic variations of AD with FDG-PET might have pathogenic and prognostic implications that can be used for personalized treatment approaches. The judicious use of FDG-PET is needed to accelerate the evaluation of promising new drugs and more rationally target treatments for dementing diseases.