The studies presented in this report were designed to assess the correlation of the bovine brain microvessel endothelial cell (BBMEC) apparent permeability coefficient (P(app)) and in vivo BBB penetration using microdialysis sampling. A mathematical model was developed to describe the relationship of brain extracellular fluid (ECF) concentration to free drug in plasma. The compounds studied have a broad range of physico-chemical characteristics and have widely varying in vitro and in vivo permeability across the blood-brain barrier (BBB). BBMEC permeability coefficients vary in magnitude from a low of 0.9 x 10(-5) cm/s to a high value of 7.5 x 10(-5) cm/s. Corresponding in vivo measurements of BBB permeability are represented by clearance (CL(in)) into the brain ECF and range from a low of 0.023 microl/min/g to a high of 12.9 microl/min/g. While it is apparent that in vitro data from the BBMEC model can be predictive of the in vivo permeability of a compound across the BBB, there are numerous factors both prior to and following entry into the brain which impact the ultimate uptake of a compound. Even in the presence of high BBB permeability, factors such as high plasma protein binding, active efflux across the BBB, and metabolism within the CNS can greatly limit the ultimate concentrations achieved. In addition, concentrations in the intracellular space may not be the same as concentrations in the extracellular space. While these data show that the BBMEC permeability is predictive of the in vivo BBB permeability, the complexity of the living system makes prediction of brain concentrations difficult, based solely on the in vitro measurement.