This article provides an overview of methods for assessing the reliability and uncertainty of quantitative structure-activity relationship (QSAR) models, particularly in the context of regulatory acceptance for human health and environmental applications. It highlights useful diagnostic tools and data analytical approaches, focusing on defining the applicability domain of a QSAR and estimating parameter and prediction uncertainty. The article discusses the importance of rigorous and independent validation of QSARs for regulatory acceptance and implementation. Key topics include the role of pattern recognition in QSARs, the selection of training sets, the demands on chemical and biological data, and various modeling techniques such as multiple linear regression (MLR), partial least squares (PLS), and other multivariate projection methods. The article also covers preprocessing techniques, informative model parameters, and methods for assessing predictive power, including cross-validation and response permutation testing. Finally, it provides recommendations for acceptability criteria and emphasizes the need for a comprehensive approach to QSAR validation.This article provides an overview of methods for assessing the reliability and uncertainty of quantitative structure-activity relationship (QSAR) models, particularly in the context of regulatory acceptance for human health and environmental applications. It highlights useful diagnostic tools and data analytical approaches, focusing on defining the applicability domain of a QSAR and estimating parameter and prediction uncertainty. The article discusses the importance of rigorous and independent validation of QSARs for regulatory acceptance and implementation. Key topics include the role of pattern recognition in QSARs, the selection of training sets, the demands on chemical and biological data, and various modeling techniques such as multiple linear regression (MLR), partial least squares (PLS), and other multivariate projection methods. The article also covers preprocessing techniques, informative model parameters, and methods for assessing predictive power, including cross-validation and response permutation testing. Finally, it provides recommendations for acceptability criteria and emphasizes the need for a comprehensive approach to QSAR validation.