Mechanism-based PKPD models are based on principles from systems biology and contain specific expressions to characterize processes on the causal path between plasma concentration and response. This includes a) the target distribution, b) the target interaction/activation and c) the transduction and homestatic control mechanisms, which may be operative. The utilisation of these models relies on novel biomarkers characterising specific processes on the causal path in a quantitative manner. An essential feature of mechanism-based PKPD models is the strict distinction between "drug-specific" and "biological system-specific" pharmacodynamic parameters to describe in vivo drug effects. We have successfully developed mechanism-based PKPD models for drugs acting at various targets including A1 Adenosine, m Opioid, 5-HT1A Serotonin and GABAA receptors. Our findings show that in general a drug's in vivo intrinsic efficacy can be accurately predicted on the basis of in vitro bioassays. Prediction of the in vivo potency on the other hand appears to be more difficult, presumably as result of complexities at the level of the target site distribution. Our results also show that equilibrium concentration-effect relationships can be readily scaled from pre-clinical animal models to humans. In contrast, the scaling of transduction and homeostatic feedback mechanisms appears to be more complex. It is concluded that mechanism-based PKPD models provide a scientific basis for the prediction of efficacy and safety of novel drugs in humans on the basis information from in vitro bioassays and/or in vivo animal studies.