Previous publications suggest that interstitial fluid compound concentrations (C(ISF)) best determine quantitative neurotherapeutic pharmacology relationships, although confirming large animal C(ISF) remains elusive. Therefore, this work primarily evaluated using respective acute dose, rat-derived unbound brain compound concentration-to-unbound plasma compound concentration ratios (C(b,u)/C(p,u)) to project accurately dog and nonhuman primate (nhp) C(b,u), a C(ISF) surrogate, from measured C(p,u) for the highly permeable non-P-glycoprotein substrates N-{(3R,4S)-3-[4-(5-cyano-2-thienyl)phenyl]tetrahydro-2H-pyran-4-yl}propane-2-sulfonamide (PF-4778574) and [4-chloro-5-fluoro-2-(3-methoxy-2-methyl-phenoxy)-benzyl]-methylamine (CE-157119) and the P-glycoprotein substrates risperidone and 9-hydroxyrisperidone. First, in rats, it was determined for eight of nine commercial compounds that their single-dose-derived C(b,u)/C(p,u) were ≤2.5-fold different from their steady-state values; for all nine drugs, their C(b,u)/C(p,u) were ≤2.5-fold different from their steady-state C(ISF)/C(p,u) (Drug Metab Dispos 37:787-793, 2009). Subsequently, PF-4778574, CE-157119 and risperidone underwent rat, dog, and nhp neuropharmacokinetics studies. In large animals at each measured C(p,u), the methodology adequately predicted [estimated mean (95% confidence interval) of 1.02 (0.80, 1.29)] the observed C(b,u) for PF-4778574 and CE-157119 but underpredicted [0.17 (0.12, 0.22)] C(b,u) for risperidone and 9-hydroxyrisperidone. The data imply that forecasting higher species C(b,u) from a measured C(p,u) and rat acute dose-determined C(b,u):C(p,u) is of high confidence for nonefflux transporter substrates that show net passive diffusion (PF-4778574) or net active influx (CE-157119) at the blood-brain barrier in rats. However, this methodology appears ineffective for correctly predicting large animal C(b,u) for P-glycoprotein substrates (risperidone and 9-hydroxyrisperidone) because of their apparently much greater C(p,u)-favoring C(b,u):C(p,u) asymmetry in rats versus dogs or nhp. Instead, for such P-glycoprotein substrates, large animal-specific cerebrospinal fluid compound concentrations (C(CSF)) seemingly best represent C(b,u).