Predictive, cost effective and fast evaluation of CNS penetration or unbound brain levels of NCEs, or both, presents a considerable challenge in preclinical screening in the pharmaceutical industry. It is evident that neither a single in vitro model nor a set of physicochemical parameters is capable of predicting complex in vivo measures, such as brain distribution or blood-brain barrier (BBB) penetration, accurately. Various in vitro and in silico strategies have been developed to model drug action as a function of BBB properties, with models that provide sufficient predictivity, reliability and throughput to be favored by the pharmaceutical industry. This review gives a practical summary of in vitro and in silico approaches to evaluate BBB penetration.