Drug-induced torsades de pointes (TdP) is a rare, but potentially lethal, unwanted effect of drugs, including many commonly prescribed noncardiac drugs. Despite its low frequency, drug-induced TdP has generated a great deal of angst among physicians and pharmaceutical companies as well as tragedy, albeit rare, among patients. Although in retrospect many patients who died suddenly as a result of drug-induced TdP had identifiable risk factors, prediction in individual cases remains problematic. Over the past decade, tremendous progress has been made with respect to elucidating the fundamental pathogenic mechanisms that underlie drug-induced TdP. The vast majority of drugs associated with "QT liability" and the potential for drug-induced TdP, including all of the drugs removed from the market because of this side effect, are "HERG (human ether-á-go-go-related gene) blockers." These drugs inhibit the KCNH2-encoded HERG potassium channel, which is one of the critical repolarizing forces involved in the exquisite orchestration of the heart's action potential. Consequently, myocyte repolarization is potentially delayed as evidenced by prolongation of the QT interval, thus providing the substrate for drug-induced TdP. Rare mutations in KCNH2 provide the pathogenic substrate for type 2 congenital long QT syndrome (LQTS), thus placing this cardiac potassium channel squarely in the intersection between congenital LQTS (the "Rosetta stone" of the heritable channelopathies) and acquired LQTS (drug-induced TdP). In addition, common polymorphisms residing in the LQTS-causing channel genes may confer heightened arrhythmogenic susceptibility and contribute to the makings of a vulnerable host. This review focuses on the present strategy of identifying "at-risk compounds" and the potential future strategy involving pharmacogenetics to pinpoint "at-risk hosts" in an effort to curb this rare, unintended, but potentially life-threatening side effect.