Summary: To further our understanding of the best way to measure regional CBF with positron emission tomog raphy (PET), we directly compared two candidate tracers W50]water and e IC]butanol, administered intrave nously) and two popular implementations of the one compartment (IC) model: the autoradiographic imple mentation representing a single PET measurement of tissue radioactivity over I min and a dynamic implemen tation representing a sequence of measurements of tissue radioactivity over 200 s. We also examined the feasibility of implementing a more realistic, and thus more complex, distributed-parameter (DP) model by assigning fixed val ues for all of its parameters other than CBF and tracer volume of distribution (V d)' a requirement imposed by the low temporal resolution and statistical quality of PET data. The studies were performed in three normal adult human subjects during paired rest and visual stimulation. In each subject seven regions of interest (ROIs) were seVarious methods for measuring regional CBF (rCBF) with positron emission tomography (PET) have been proposed, both for functional brain map ping and for determination of cerebral metabolic supply-demand relationships. Many of these have employed e 5 0]water as the tracer, primarily due to the short half-life of 1 5 0, which allows sequential studies and exposes the subject to a relatively low radiation dose. Although the methods of tracer ad ministration vary , including continuous inhalation of e 5 0]carbon dioxide (Frackowiak et aI. , 1980; Brooks et aI. , 1986; Lammertsma et aI. , 1989), sin gle-breath inhalation of e 5 0]carbon dioxide (Kanno et aI. , 1984), ramp intravenous infusion of [1 5 0]wa_ ter (Ginsberg et aI. , 1982), and intravenous bolus injection of e 5 0]water (Herscovitch et aI. , 1983; Huang et aI. , 1983 ; Raichle et aI. , 1983), most anal yses of the resultant data are based on a one-