(R)-2-Amino-1,3',3'-trimethyl-7'-(pyrimidin-5-yl)-3',4'-dihydro-2'H-spiro[imidazole-4,1'-naphthalen]-5(1H)-one (GNE-892) is an orally administered inhibitor of β-secretase 1 (β-site amyloid precursor protein cleaving enzyme 1, BACE1) that was developed as an intervention therapy against Alzheimer's disease. A clinical microdosing strategy was being considered for de-risking the potential pharmacokinetic liabilities of GNE-892. We tested whether dose-proportionality was observed in cynomolgus monkey as proof-of-concept for a human microdosing study. With cryopreserved monkey hepatocytes, concentration-dependency for substrate turnover and the relative contribution of P450- versus AO-mediated metabolism were observed. Characterization of the kinetics of these metabolic pathways demonstrated differences in the affinities of P450 and AO for GNE-892, which supported the metabolic profiles that had been obtained. To test if this metabolic shift occurred in vivo, mass balance studies in monkeys were conducted at doses of 0.085 and 15 mg/kg. Plasma exposure of GNE-892 following oral administration was more than 20-fold greater than dose proportional at the high-dose. P-gp-mediated efflux was unable to explain the discrepancy. The profiles of metabolites in circulation and excreta were indicative that oxidative metabolism limited the exposure to unchanged GNE-892 at the low dose. Further, the in vivo data supported the concentration-dependent metabolic shift between P450 and AO. In conclusion, microdosing of GNE-892 was not predictive of pharmacokinetics at a more pharmacologically relevant dose due to saturable absorption and metabolism. Therefore, it is important to consider ADME liabilities and their potential concentration-dependency when deciding upon a clinical microdosing strategy.