Structure-based drug design seeks to exploit the structure of protein-ligand or protein-protein binding sites, but the site is not always known at the outset. Even when the site is known, the researcher may wish to identify alternative prospective binding sites that may result in different biological effects or new class of compounds. It is also vital in lead optimization to clearly understand the degree to which known binders or docking hits satisfy or violate complementarity to the receptor. SiteMap is a new technique for identifying potential binding sites and for predicting their druggability in lead-discovery applications and for characterizing binding sites and critically assessing prospective ligands in lead-optimization applications. In large-scale validation tests, SiteMap correctly identifies the known binding site in > 96% of the cases, with best results (> 98%) coming for sites that bind ligands tightly. It also accurately distinguishes between sites that bind ligands and sites that don't. In binding-site analysis, SiteMap provides a wealth of quantitative and graphical information that can help guide efforts to modify ligand structure to enhance potency or improve physical properties. These attributes allow SiteMap to nicely complement techniques such as docking and computational lead optimization in structure-base drug design.