Canine cranial cruciate ligament rupture (CCLR) is the most common cause of pelvic limb lameness in dogs. CCLR results in palpable stifle instability secondary to cranial translation of the tibial relative to the femur, and it can be diagnosed during the orthopedic exam using the cranial drawer test (CDT) and tibial compression test (TCT). Accurate diagnosis of CCLR depends on the efficacy in performing these tests. In this study, two three-dimensional canine pelvic limb models were developed: one simulating a normal stifle and one simulating CCLR. Thirty-eight veterinary student participants answered questionnaires and performed both the CDT and TCT on a randomly assigned model. Twenty-one participants also manipulated the models one week later to assess skill retention in the short term. Mean levels of reported confidence in diagnosing CCLR and finding anatomic landmarks for CDT/TCT were significantly higher following model manipulation. Nearly all participants reported that they desired a model for teaching the diagnosis of CCLR. Most participants (92.5%) felt that the tested model would be useful for teaching CCLR diagnosis. Accuracy in diagnosing CCLR with the TCT significantly improved over time. Participant response indicated that while the tested model was effective and desirable, an ideal model would be more durable and lifelike. Further studies are needed to evaluate the developed models' effectiveness for teaching CCLR diagnosis compared to traditional teaching methods.