Recurrent acute and/or chronic stress can affect all vertebrate species, and can have serious consequences. It is increasingly and widely appreciated that laboratory animals experience significant and repeated stress, which is unavoidable and is caused by many aspects of laboratory life, such as captivity, transport, noise, handling, restraint and other procedures, as well as the experimental procedures applied to them. Such stress is difficult to mitigate, and lack of significant desensitisation/habituation can result in considerable psychological and physiological welfare problems, which are mediated by the activation of various neuroendocrine networks that have numerous and pervasive effects. Psychological damage can be reflected in stereotypical behaviours, including repetitive pacing and circling, and even self-harm. Physical consequences include adverse effects on immune function, inflammatory responses, metabolism, and disease susceptibility and progression. Further, some of these effects are epigenetic, and are therefore potentially transgenerational: the biology of animals whose parents/grandparents were wild-caught and/or have experienced chronic stress in laboratories could be altered, as compared to free-living individuals. It is argued that these effects must have consequences for the reliability of experimental data and their extrapolation to humans, and this may not be recognised sufficiently among those who use animals in experiments.