Introduction: Neurotoxicity is a critical liability for many environmental pollutants. Current in vitro neurotoxicity screens rely on direct exposure of cultured neurons to xenobiotics, often at exceeding physiologically relevant levels due to the restrictive nature of the blood -brain barrier (BBB). This limitation reduces the accuracy of central nervous system (CNS) exposure predictions. Methods: To address this limitation, we have developed a novel human in vitro direct-contact triculture BBB model that more closely mimics the in vivo barrier. The triculture is formed by layering primary astrocytes, primary pericytes, and then brain microvessel endothelial cells (BMECs, HBEC-5i) in direct contact, increasing the restrictive nature of tight junctions and allowing cell -cell signaling that mimics the configuration found in the in vivo BBB. Using this model, we quantified the apparent bidirectional permeability (P Results: The triculture model demonstrated enhanced tight junction organization and increased efflux transporter expression compared with endothelial monocultures, indicating an improved barrier phenotype. Using our measured bidirectional Papp values, calculated efflux ratios, and EPA physiologically based pharmacokinetic (PBTK) reference data for compound parameters, we are developing predictions of toxicant accumulation in the brain parenchyma after chronic exposure in steady state. Discussion: Integration of