Editorial COMPUTER-AIDED D R U G DISCOVERY AND DESIGN The impact of pharmaceuticals on the human condition is undeniable. Equally clear is the dramatic cost of drug discovery and development. These costs are measured in years, sometimes in decades, and in hundreds of millions of dollars. In an attempt to streamline the drug discovery process, a variety of computational tools have been developed and applied to a wide range of projects in industry and at universities. This volume begins with Topliss' historical discussion of quantitative structure-activity relationships (QSAR), an extremely practical approach to the optimization of an often serendipitous discovery. QSAR exploits a simple view of ligand structure versus activity in the conventional setting, where little is known about the three-dimensional organization of a receptor or enzyme. With ever-expanding libraries of small molecules, this analysis naturally migrated from medicinal chemists' notes to computers, to take advantage of the graphics, speed and seemingly boundless memory of such devices, as reviewed in this volume by Boyd and Milosevich, Olson and Morris, and Spellmeyer and Swope. QSAR has thus become a standard tool among the many armaments of pharmaceutical scientists. Recent attempts to expand the utility of this strategy are reviewed for example by Cramer, King, Hirst and Sternberg, and by Crippen and colleagues. With advances in methods to determine or hypothesize three-dimensional structures of pharmaceutically relevant targets, the field of structure-based drug design has become ever more prominent, as illustrated by Kontoyianni and Lybrand and by Zuiderweg and colleagues. This is truly an exciting area, with significant potential to drive lead compound discovery and subsequent optimization. The approach is further highlighted in the discussions by Blaney and Dixon, and Reich and Webber. Many compounds currently in clinical trials have major or at least partial origins in the structure-based approaches. While much has been accomplished with computer-aided discovery and structure-based design, much work remains. In parallel, evolving discovery techniques that allow the combinatorial generation and screening of vast libraries of molecular diversity have renewed interest in empirical brute-force strategies. Nonetheless, for appropriately chosen problems, by itself and in support of other approaches, drug design can play a preferred role in speeding the search for new pharmaceuticals with specific activities. However, the implication of this work for other, equally important drug development areas, such as pharmacokinetics and toxicology, remains to be demonstrated. Fred E. Cohen Walter Moos