Endothelial cells forming the blood-brain barrier limit drug access into the brain, due to tight junctions, membrane drug transporters, and unique lipid composition. Passive permeability, thought to mediate drug access, is typically tested using porcine whole-brain lipid. However, human endothelial cell lipid composition differs. This investigation evaluated the influence of lipid composition on passive permeability across artificial membranes. Permeability of CNS-active drugs across an immobilized lipid membrane was determined using three lipid models: crude extract from whole pig brain, human brain microvessel lipid, and microvessel lipid plus cholesterol. Lipids were immobilized on polyvinylidene difluoride, forming donor and receiver chambers, in which drug concentrations were measured after 2 h. The log of effective permeability was then calculated using the measured concentrations. Permeability of small, neutral compounds was unaffected by lipid composition. Several structurally diverse drugs were highly permeable in porcine whole-brain lipid but one to two orders of magnitude less permeable across human brain endothelial cell lipid. Inclusion of cholesterol had the greatest influence on bulky amphipathic compounds such as glucuronide conjugates. Lipid composition markedly influences passive permeability. This was most apparent for charged or bulky compounds. These results demonstrate the importance of using species-specific lipid models in passive permeability assays.