J. J. Runge 1 , S. P. Kelly 1 , T. P. Gregor 1 , S. Kotwal 1 and G. K. Smith 1 1 Department of Clinical Studies, Matthew J. Ryan Veterinary Hospital at the University of Pennsylvania, 3900 Delancey Street, Philadelphia, PA, USA JSAP May 2010 51 5 264-269 ABSTRACT Objective: To determine if age, breed, gender, weight or distraction index (DI) influenced the risk of radiographic osteoarthritis (OA) of canine hip dysplasia (CHD) in four common dog breeds; the American bulldog, Bernese mountain dog, Newfoundland and standard poodle. Materials and Methods: This was a cross sectional prevalence study with 4349 dogs. Canine hips were evaluated using 3 radiographic projections: the hip-extended view, the compression view and the distraction view. The hip-extended view was examined for the presence of OA. The PennHIP distraction view was utilized to calculate the DI. For all breeds, a multiple logistic regression model incorporating age, weight, gender, and DI was created. For each breed, disease-susceptibility curves grouping dogs on the basis of age were constructed. Receiver-operating characteristic (ROC) curves were developed for each breed regardless of age. Results: For all breeds, DI was the most significant risk factor for the development of OA associated with CHD. Weight and age were also significant risk factors in all four breeds, but gender was not. Clinical Significance: Results from this study support previous findings, that irrespective of breed, the probability of radiographic OA increases with hip joint laxity as measured by the DI. Breed-specific differences in this relationship, however, warrant investigation of all breeds affected by CHD to determine inherent dependency of hip OA on joint laxity. Such findings guide veterinarians in helping dog breeders to make evidence-based breeding decisions and in informing dog owners to implement preventative treatments for CHD for dogs found to be at risk.