-Animal models of human disease have been widely used in drug discovery, but they are rarely utilized in toxicological research and screening (except for transgenic models in carcinogenicity testing). Although genetic and/or acquired pathophysiological alterations associated with a particular disease may greatly exacerbate toxic responses to drugs in certain patient subsets, these pre-existing pathological conditions are usually not considered in preclinical safety assessment. Examples of disease-related determinants of susceptibility include disruption of the cytokine network in pro-inflammatory conditions, mitochondrial alterations and oxidative stress in certain neurodegenerative diseases, altered antioxidant defense in certain viral infections, and altered gene expression and mitochondrial dysfunction in type 2 diabetes. Hence, if cellular stress caused by drugs or metabolites and the disease-related effects are superimposed, then an individual can become sensitized to potential drug toxicity. Animal models of modest inflammation indeed can potentiate the toxicity of certain drugs. Similarly, rodent models of type 2 diabetes predispose the animals to hepatotoxic effects of thiazolidinediones antidiabetics. In conclusion, it is suggested that tailor-made and simplified models be adopted and increasingly used, in spite of clear limitations, as optimal substrates for satellite toxicity studies to facilitate candidate selection, help predict rare and unexpected toxicity, and identify new biomarkers.