Developments in biomedical science continue to transform our understanding of concepts such as health and disease. The creation of this expertise has also had a substantive role in changing the veterinary approach to animal diseases. Traditionally, companion animal veterinarians modelled their practices on developments in the diagnosis and treatment of human patients. As science and technology have realigned the boundaries between normalcy, intra-species variation and pathology in particular domains of expertise such as orthopaedic surgery, these patterns of knowledge translation have changed. Not so long ago, treatments for the rupture of the cruciate ligament in human and canine patients were based on pathoanatomical comparison and designed to reestablish the stability of the joint by the functional restoration of ligament anatomy. Recently, a radically different characterization of the canine injury-with a corresponding alternative intervention-has been proposed within the field of veterinary orthopaedics. It views the normal anatomy of the canine knee as being in some way inherently pathogenic and proposes the surgical creation of an idealized structure as a remedy to dysfunction and disease. The veterinary focus on an ideal of patient performance, rather than a specific pathology, is now influencing how orthopaedists choose to approach analogous human injuries. In this article, I chart the history of canine 'cruciate disease' therapies as a means to map some of the epistemic assumptions, interplays of idealized and analogical reasoning and patterns of knowledge translation central to the biomedicalization of the science and practice of orthopaedic surgery. © 2010 Springer Science+Business Media B.V.