OBJECTIVE To compare the accuracy of ultrasonography and MRI for diagnosing medial meniscal lesions in dogs with cranial cruciate ligament (CCL) deficiency. DESIGN Diagnostic test evaluation. ANIMALS 26 dogs (31 stifle joints) with CCL deficiency. PROCEDURES A single surgeon physically examined each dog and performed ultrasonography and arthroscopy of affected stifle joints to identify medial meniscal lesions. Video recordings of the arthroscopic procedure were saved and subsequently reviewed by the same surgeon and by a second surgeon working independently and blinded to results of all examinations. A radiologist blinded to results of all examinations evaluated MRI scans of the affected joints. Correct classification rate (CCR), sensitivity, and specificity of ultrasonography and MRI were calculated twice, with each of the 2 surgeons' arthroscopic assessments used as the reference standard. RESULTS Compared with arthroscopic examination by the unblinded surgeon, ultrasonography had a CCR of 90%, sensitivity of 95% (95% confidence interval [CI], 73% to 100%), and specificity of 82% (95% CI, 48% to 97%). For MRI, these values were 84%, 75% (51% to 90%), and 100% (68% to 100%), respectively. Compared with arthroscopic assessment by the blinded surgeon, ultrasonography had a CCR of 84%, sensitivity of 86% (95% CI, 64% to 96%), and specificity of 78% (95% CI, 40% to 96%). For MRI, these values were 77%, 68% (45% to 82%), and 100% (63% to 100%), respectively. CONCLUSIONS AND CLINICAL RELEVANCE These data suggested imperfect performance but clinical usefulness of both ultrasonography and MRI for diagnosing medial meniscal lesions in dogs.