The aim of this study was to test the hypothesis that some cases of drug-induced arrhythmias depend on genetic predisposition. Excessive prolongation of the QT interval and life-threatening arrhythmias (torsades de pointes or ventricular fibrillation) may occur in response to a variety of cardiac and noncardiac drugs, with detrimental effects on patient safety and the investments made by the pharmaceutical industry. Moss and Schwartz hypothesized that some drug-induced arrhythmias might represent cases of "forme fruste" of the congenital long QT syndrome (LQTS). The availability of molecular screening techniques for LQTS genes allowed us to test this hypothesis. An elderly female patient with documented cardiac arrest related to cisapride, a prokynetic drug that blocks I(Kr), and transiently prolonged QT interval underwent mutational analysis of the known LQTS-related genes performed by single-strand conformational polymorphism and DNA sequencing. Double-electrode voltage clamp in Xenopus oocytes as the expression system was used to study the in vitro cellular phenotype caused by the genetic defect in coexpression with the wild-type (WT) gene. Molecular analysis revealed a heterozygous mutation leading to substitution of a highly conserved amino acid in the pore region of KvLQT1. This mutation was present not only in the patient with ventricular fibrillation but also in her two adult asymptomatic sons who have a normal QT interval. In vitro expression of the mutated KvLQT1 protein showed a severe loss of current with a dominant negative effect on the WT-KvLQT1 channel. Our findings demonstrate that some cases of drug-induced QT prolongation may depend on a genetic substrate. Molecular screening may allow identification among family members of gene carriers potentially at risk if treated with I(Kr) blockers. Evolving technology may lead to rapid screening for mutations of candidate genes that cause drug-induced life-threatening arrhythmias and allow early identification of individuals at risk.