It is becoming increasingly clear that the interactions of targets and biomarkers, drug modes of action and molecular mechanisms of side-effects and toxic effects are much more complex than previously anticipated, basically due to physiological compensation and cross-talk. Single genes often lead to hundreds or even thousands of functional protein molecules, modified at the post-translational level. Thus, the comprehensive analysis of proteins (proteomics) teaches us that physiological activity means dynamic, multidimensional processes among many thousands of different proteins within higher systems of organisation and correlation. Crucial for control and relevant reduction of this enormous complexity, which will enable new kinds of molecular drug screening, as well as a new type of molecular toxicology, is a consequently differential and quantitative protein analysis. Precise knowledge of key protein isoforms with specific post-translational modifications within kinetic and contextual relationships is accessible by powerful new technologies, which have emerged to analyse the surprisingly ambiguous world of proteins, where single molecular modules are involved in a diversity of often opposing signal transduction pathways in a most flexible way.