Formation of reactive intermediates by metabolism of xenobiotics represents a potential liability in drug discovery and development. Although it is difficult, if not impossible, to predict toxicities of drug candidates accurately, it is prudent to try to minimize bioactivation liabilities as early as possible in the stage of drug discovery and lead optimization. Measurement of covalent binding to liver microsomal proteins in the presence and the absence of NADPH, as well as the use of trapping agents such as glutathione or cyanide ions to provide structural information on reactive intermediates, have been used routinely to screen drug candidates. These in vitro experiments are often supplemented with in vivo covalent binding data in rats. The resulting data are not only used to eliminate potentially risky compounds, but, more importantly, they provide invaluable information to direct the Medicinal Chemistry group efforts to design analogs with less propensity to undergo bioactivation. Select case histories are presented in which this approach was successfully applied at Merck.