High-throughput screening (HTS) for central nervous system (CNS) agents has emerged as a driving force in the identification of potential novel lead structures. Various virtual screening techniques complement these efforts. Computer-assisted classi-fication and visualization of HTS data, similarity searching routines, pharmacophore modeling, the rational shaping of compound libraries, and de novo design suggestions can provide useful information for many medicinal chemistry projects. Virtual screening methods have been successfully applied to generating high-quality screening libraries, and identifying novel molecular scaffolds that can serve as “fast followers” and thus lead to competitive advantage, or “second series structures” with improved pharmacokinetic properties. In CNS-relevant drug development the main targets are integral membrane proteins, in particular G-protein coupled receptors. A major challenge lies in the extraction of relevant, function-determining molecular features from HTS hits or from only a single known active compound in the absence of a high-resolution receptor structure. One aim is to identify novel molecular architectures exhibiting substantial bioactivity that is comparable to the activity of previously known ligands. Virtual screening can help complement the random HTS approach by a more rational, model-based molecular design and selection process. This review article gives an overview of current virtual screening concepts along with several applications. Special emphasis is put on orphan receptor (ORL-1) ligand screening, and the search for calcium and potassium ion channel-blocking agents.