We review our computational tools for high-throughput screening by fragment-based docking of large collections of small molecules. Applications to six different enzymes, four proteases, and two protein kinases, are presented. Remarkably, several low-micromolar inhibitors were discovered in each of the high-throughput docking campaigns. Probable reasons for the lack of submicromolar inhibitors are the tiny fraction of chemical space covered by the libraries of available compounds, as well as the approximations in the methods employed for scoring, and the use of a rigid conformation of the target protein.