The ability to predict clinical safety based on chemical structures is becoming an increasingly important part of regulatory decision making. (Quantitative) structure-activity relationship ((Q)SAR) models are currently used to evaluate late-arising safety concerns and possible nonclinical effects of a drug and its related compounds when adequate safety data are absent or equivocal. Regulatory use will likely increase with the standardization of analytical approaches, more complete and reliable data collection methods, and a better understanding of toxicity mechanisms.