In drug development, human predictability of non-clinical pharmacological evaluations should be improved for high success probability of clinical studies. In vitro in vivo extrapolation (IVIVE) is key methodology for this purpose. IVIVE has been eagerly studied to predict pharmacokinetics (PK) and pharmacodynamics (PD), thereby determining the therapeutic window of a drug. In the central nervous system (CNS), to examine whether or not candidate compounds could penetrate across the human blood-brain barrier (BBB), a variety of in vitro BBB models have been reported. On the other hand, in the CNS, specific regions are responsible for specific brain functions including higher-order functions. That means where and what candidate compounds are doing in the CNS is important to predict their efficacies and adverse effects. This information will also contribute to more accurate prediction of the therapeutic window. Currently we are focusing on in vivo imaging systems such as MRI, etc.. Now human neurons, such as human induced pluripotent stem cell-derived neurons (hiPSC-neurons), can be used in in vitro pharmacological evaluations. The human CNS predictability of non-clinical pharmacological evaluations is expected to be improved by unifying the multiphasic data described above.