Articular cartilage defects are one of the features of osteoarthritis in animals and humans. Early detection of cartilage defects is a challenge in clinical veterinary practice and also in translational research studies. An accurate, diagnostic imaging method would be desirable for detecting and following up lesions in specific anatomical regions of the articular surface. The current prospective experimental study aimed to describe the accuracy of computed tomographic arthrography (CTA) for detecting cartilage defects in a common animal model used for osteoarthritis research, the ovine stifle (knee, femoropatellar/femorotibial) joint. Joints in cadaver limbs (n = 42) and in living animals under anesthesia (n = 13) were injected with a contrast medium and imaged using a standardized CT protocol. Gross anatomy and histological assessment of specific anatomic regions were used as a gold standard for the evaluation of sensitivity, specificity, negative predictive value, and positive predictive value for CTA identification of articular cartilage defects in those regions. Pooled estimated sensitivity and specificity were 90.32% and 97.30%, respectively, in cadaver limbs, and 81.82% and 95.24%, respectively, in living animals. Pooled estimated positive predictive value and negative predictive values were 98.25% and 85.71%, respectively, in cadaver limbs, and 81.82% and 95.24%, respectively, in living animals. The delineation of cartilage surface was good for anatomical regions most frequently affected by cartilage defects in the ovine stifle: medial femoral condyle, medial tibial condyle, and patella. This study supported the use of CTA as an imaging technique for detecting and monitoring articular cartilage defects in the ovine stifle joint.