[en] Complete text of publication follows: Objectives: P-glycoprotein (ABCB1) is the most studied ATP-binding cassette (ABC) transporter expressed at the blood-brain barrier (BBB). A powerful method to study P-gp function at the BBB is Positron Emission Tomography (PET) imaging in combination with radiolabeled P-gp substrate radiotracers. Currently available radiotracers are high-affinity P-gp substrates and convincingly highlight this transporter as a functional component of the BBB, which restricts the influx of its substrates from blood to the brain. Many CNS-active drugs, such as metoclopramide, were shown to be weak P-gp substrates and nonetheless show sufficient permeability to cross the BBB and exert CNS effects. The impact of P-gp on the brain kinetics of such weak substrates is not known. Methods We have recently shown the feasibility of PET imaging using ¹¹C-metoclopramide to measure in vivo P-gp function at the rodent BBB. ¹¹C-metoclopramide showed good specificity towards P-gp, suitable radiochemical purity of the PET signal in the brain (absence of radiometabolites in the brain) and lack of specific binding to other target structures in the brain. In the present study, ¹¹C-metoclopramide PET imaging was performed in non-human primates, a relevant animal model of the human BBB in terms of P-gp expression. Brain ¹¹C-metoclopramide (298.12 ± 44.13 MBq) PET acquisitions (60 min), including the measurement of the metabolite-corrected arterial input function, were performed in 4 anesthetized baboons (propofol). Experiments were performed in each animal in the absence (baseline) and the presence of an i.v infusion protocol allowing for constant and controlled plasma concentrations of the potent P-gp inhibitor tariquidar (TQD). PET images were co-registered onto corresponding MR images for each baboon to generate corresponding time-activity curves in selected brain regions. Kinetic modeling was performed using a 1-tissue compartment model (1-TCM) to estimate the influx (K1; mL/min/cm³) and efflux (k2; min^-1) rate constants and the total volume of distribution (VT = K1/k2 in the 1-TCM; mL/cm³). The wash-out of ¹¹C-metoclopramide from the brain was also described by the elimination slope kE (min-1), graphically measured from the Log-transformed PET kinetics from 30 to 60 min. Outcome parameters were statistically compared using a 2-way ANOVA with 'treatment' and 'brain regions' as factors. Results Baseline PET images showed the brain distribution of ¹¹C-metoclopramide-associated radioactivity, which was homogeneously distributed among brain regions. TQD increased the brain uptake of ¹¹C-metoclopramide and notably decreased its wash-out from the brain with a significant 1.5±0.2-fold decrease in kE (p≤0.01). PET kinetics were accurately estimated using the 1-TC model. Compared with baseline (VT = 4.3±0.5 mL.cM^-3), TQD significantly increased the brain distribution of ¹¹C-metoclopramide (VT = 8.7±0.5 mL.cM^-3, p≤0.001). This increase in VT was due to a significant 1.3±0.1-fold increase in K1 (p≤0.05) and a 1.6 ± 0.1-fold decrease in k2 (p≤0.001). The effect was homogeneous across tested brain regions. Conclusion Using ¹¹C-metoclopramide PET imaging in baboons, we showed that P-gp does not solely act as a 'barrier' to limit the brain penetration of xenobiotics from the blood. P-gp also mediates the clearance of its substrates back to the blood, thus providing an additional dynamic system to limit overall brain exposure