The use of genetically manipulated animals in conjunction with classical physiological and biochemical measurement has unravelled many pathological changes in animal models of chronic pain that bear some striking similarities to those described in several chronic pain conditions in humans. In this article, I highlight several limitations in the validation of animal models of chronic pain and the methods that are used for assessing pain-like behaviours in these models. Alternative methods for assessing pain and stress in animals, which might better reflect the diverse symptomotology of chronic pain in humans, are proposed.