The applicability of using the standard 3-compartment model to describe the neuropharmacokinetics of a high affinity substituted benzamide was investigated. We performed the following experiments using the [18F]-5-(3-fluoropropyl) analog of epidepride ([18F]5-FPrEpid), a potent dopamine D2 receptor antagonist: constant left ventricular infusion, first-pass clearance, varying ligand specific activity, and displacing bound ligand with varying amounts of unlabelled ligand. Taken together, the information from these experiments rigorously tests the standard 3-compartment model. The obtained data and predictions from the model of the kinetic behavior of the ligand are inconsistent. The measured and model predicted dissociation rate (measured koff = 0.065 min-1, model prediction koff = 0.007 min-1) and the equilibrium dissociation constant (measured KD = 0.14 nM, model prediction KD = 2.2 nM) differ by an order of magnitude. Furthermore, the model cannot be used to accurately estimate the receptor density. We postulate that the synapse geometry and physical relationship between receptors are necessary components of a model that describes the pharmacokinetics of [18F]5-FPrEpid.