This paper examines evidence which suggests that the self-mutilation of deafferented limbs exhibited by laboratory animals is a response to pain or dysesthesia and is therefore an adequate model of chronic pain. Evidence from studies using physiological, pharmacological and behavioral methods provides strong support that autotomy reflects chronic pain. New evidence presented in this paper demonstrates that specific treatments can be used to manipulate the extent of autotomy, causing increases or decreases, as well as restricting it to specific parts of a denervated foot. This evidence argues that autotomy scores are an appropriate measure of the degree of pain or dysesthesia which results from the deafferentation of a limb.