KiBank is a database for computer-aided drug design and consists of binding affinities and chemical and target protein structures. Each chemical or protein structure with hydrogen atoms added was optimized by energy minimization and stored in PDB or MDL MOL file format, so that the structural data can be directly used for in silico binding studies. To describe the extent of inhibition, the inhibition constant (K(i)) value is used to simplify comparisons of strengths among chemical-protein bindings. As of April 2004, KiBank contained 142 proteins, over 5000 chemicals, and over 6000 binding affinity values that were published in peer-reviewed journals. The binding affinity values are currently mostly for membrane and nuclear receptors but are soon being expanded to other drug targets. KiBank is updated daily and can be accessed on the Web at http://kibank.iis.u-tokyo.ac.jp/at no charge.