Active-site concentrations can be defined as the concentrations of unbound, pharmacologically active substances at the site of action. In contrast, the total concentrations of the drug in plasma/organ/tissue also include the protein- or tissue-bound molecules that are pharmacologically inactive. Plasma and whole tissue concentrations are used as predictors of effects and side effects because of their ease of sampling, while the concentrations of unbound drug in tissue are more difficult to measure. However, with the introduction of microdialysis and subsequently developed techniques, it has become possible to test the free drug hypothesis. The brain is an interesting organ in this regard because of the presence of the blood-brain barrier with its tight junctions and active efflux and influx transporters. We have proposed that research into brain drug delivery be divided into three main areas: the rate of delivery (PS, CL(in)), the extent of delivery (K(p,uu)) and the non-specific affinity of the drug to brain tissue, described by the volume of distribution of unbound drug in the brain (V(u,brain)). In this way, the concentration of unbound drug at the target site can be estimated from the total brain concentration and the plasma concentration after measuring the fraction of unbound drug. Results so far fully support the theory that active site concentrations are the best predictors when active transport is present. However, there is an urgent need to collect more relevant data for predicting active site concentrations in tissues with active transporters in their plasma membranes.