The article discusses the use of non-specific endpoints in traditional and alternative toxicity tests for chemical safety assessment. Non-specific endpoints, such as body weight and organ weight changes, are often used to set critical effect levels, which are then used to derive regulatory points of departure (PoD) and exposure limits. However, these non-specific endpoints may be equally or more sensitive than specific apical endpoints, such as impaired reproductive function or altered neurodevelopment. This means that animal data, which are often used to develop protective critical effect levels, may not predict specific toxicological responses in humans. The authors argue that this approach is a conservative and protective method for setting exposure limits, but it may not be as specific or predictive as desired. They suggest that a rethinking of how adverse chemical effects are interpreted is needed, and that a combination of in vitro and in silico methods could be used to more accurately discriminate between specific and non-specific effects. The article also highlights the challenges in distinguishing between systemic effects and specific toxicants in animal studies, and the need for careful data curation and controlled vocabulary to ensure accurate interpretation of results. Finally, the authors propose a conceptual framework for using new approach methods (NAMs) to detect both non-specific and specific effects, emphasizing the importance of combining protective and predictive approaches in future safety assessments.The article discusses the use of non-specific endpoints in traditional and alternative toxicity tests for chemical safety assessment. Non-specific endpoints, such as body weight and organ weight changes, are often used to set critical effect levels, which are then used to derive regulatory points of departure (PoD) and exposure limits. However, these non-specific endpoints may be equally or more sensitive than specific apical endpoints, such as impaired reproductive function or altered neurodevelopment. This means that animal data, which are often used to develop protective critical effect levels, may not predict specific toxicological responses in humans. The authors argue that this approach is a conservative and protective method for setting exposure limits, but it may not be as specific or predictive as desired. They suggest that a rethinking of how adverse chemical effects are interpreted is needed, and that a combination of in vitro and in silico methods could be used to more accurately discriminate between specific and non-specific effects. The article also highlights the challenges in distinguishing between systemic effects and specific toxicants in animal studies, and the need for careful data curation and controlled vocabulary to ensure accurate interpretation of results. Finally, the authors propose a conceptual framework for using new approach methods (NAMs) to detect both non-specific and specific effects, emphasizing the importance of combining protective and predictive approaches in future safety assessments.