Drug concentrations that would prolong repolarization parameters by 10%, including action potential duration (APD₉₀, APD_{30 – 90}), in in vitro assays using guinea-pig papillary muscle and QTc intervals in in vivo assays using conscious dogs, conscious monkeys, and anesthetized dogs were compared. Although, both the in vitro and in vivo assays showed concentration-dependent responses for compounds that have been classified as torsadogenic in humans, only a weak correlation in EC₁₀ values was observed between the in vitro and in vivo assays. Among the in vivo QT assays, the EC₁₀ values obtained from conscious dogs, conscious monkeys, and anesthetized dogs correlated well with each other, but the EC₁₀ values in monkeys were somewhat lower in comparison to those in dogs. When in vivo QT assay EC₁₀ values were compared to the respective human effective therapeutic plasma concentration (ETPC), the ratios of EC₁₀ values to ETPCs were less than 20 for most torsadogenic compounds. In conclusion, the relationships between the extent of QTc interval prolongation and the concentration of drugs was highly consistent among the three in vivo models, suggesting that the ratios of EC₁₀ values in in vivo QT assays are useful for estimating the safety margin of drugs that prolong the QTc interval.