At least two rates of dopamine turnover have been demonstrated in vivo, including a slow turnover rate that is associated with synaptic vesicles, and a faster rate that leads to rapid production of dopamine metabolites. Similarly, [18F]6-fluorodopamine (FDA), the decarboxylation product of the PET tracer [18F]6-fluoro-L-DOPA (FDOPA), may have multiple turnover rates which could substantially affect the interpretation of FDOPA uptake. To better characterize FDA turnover in vivo, we measured the formation of FDOPA metabolites in primate brain following bolus FDOPA injection with carbidopa pretreatment. FDOPA was allowed to circulate for either 30 minutes or 90 minutes, prior to removal of brain samples. The primary metabolites in striatum were [18F]6-fluoro-3-methyl-L-DOPA (3-OMFD), FDA, [18F]6-fluoro- L-3,4-dihydroxyphenylacetic acid (FDOPAC), and [18F]6-fluorohomovanillic acid (FHVA). The percentages of total radioactivity in striatum at 30 minutes and 90 minutes were: FDOPA(5%, 2%), FDA (39%, 23%), FDOPAC (12%, 3%), FHVA (14%, 34%), and 3-OMFD (29%, 39%). In cortex and cerebellum most of the activity (73%, 80%) was 3-OMFD. These data were compared against the metabolite profiles predicted by two compartmental models of FDOPA metabolism. A model that assumes only a single slow rate of FDA turnover predicted much lower concentrations of FDA metabolites (FDOPAC, FHVA) in striatum than were found in the brain assay, while a model that includes both slow and fast FDA turnover was in much better agreement. These findings extend and confirm previous observations of FDOPA metabolites. The implications for the interpretation of FDOPA PET, particularly in terms of the availability of dopamine synthesized from therapeutic L-DOPA, are discussed.