Given that the cerebral microvasculature within the brain constitutes the rate-limiting barrier to drug entry, primary cultures of cerebral capillary endothelial cells would appear to offer a potentially useful model system for predicting drug delivery to the central nervous system. In the present study, the predictive capabilities of two potential models of the in vivo blood-brain barrier (BBB) to the passive diffusion of polar permeants were assessed. A comparison of the logarithms of the in vitro transmonolayer permeability coefficients (Pmonolayer) for several polar permeants varying in lipophilicity (from this study and literature data) with the well-established relationship between the logarithms of the in vivo BBB permeability coefficients (log PBBB) and permeant lipophilicity as measured by the logarithm of the octanol/water partition coefficient (log PCoctanol/water) demonstrated that in vitro permeation across these monolayers is largely insensitive to polar permeant lipophilicity as a result of the predominance of the paracellular component in the transmonolayer flux. Conversely, kinetic studies of uptake of the same compounds into monolayers yielded transfer rate constants (kp) reflecting membrane permeability coefficients ranging over several orders of magnitude, similar to the variation in permeant lipophilicity. Furthermore, a linear relationship could be demonstrated between the logarithms of kp and in vivo BBB log P (slope = 1.42 +/- 0.35; r = 0. 92). In conclusion, this preliminary investigation suggests that monitoring the kinetics of cell uptake into cerebral capillary endothelial cell monolayers may be superior to transmonolayer flux measurements for predicting the passive diffusion of polar permeants across the BBB in vivo.