Purpose: The purpose of the study is to evaluate the usefulness of CT arthrography (CTA) by conducting a comparative study of CTA and MR arthrography (MRA) with the arthroscopic findings in the same patients. Materials and methods: Forty nine patients who suffered from shoulder disease underwent MRA and CTA concurrently. The sensitivity and specificity of CTA was evaluated, as compared to MRA, in the case of four types of shoulder pathological lesions. The accuracy of CTA was evaluated by analyzing the arthroscopic findings of thirty four patients. Results: Compared to MRA, CTA showed a sensitivity of 90% and a specificity of 100% for diagnosing supraspinatus tendon (SST) full thickness tear, and CTA showed a sensitivity of 100% and a specificity of 97.5% for making the diagnosis of superior labrum anterior to posterior (SLAP) lesion. For diagnosing partial articular side supraspinatus avulsion (PASTA) lesion, some studies have shown the usefulness of CTA with a sensitivity of 71.4% and a specificity of 97.7%. However, for diagnosing SST bursal side partial tear, the sensitivity of CTA was as low as 10%. CTA has been shown to be relatively accurate when the diagnoses were verified with the arthroscopic findings; diagnosing SST full thickness tear and SLAP lesion with CTA has shown an accuracy of 100% and87.5% respectively, and CTA showed 71.4% diagnostic accuracy for PASTA lesion. Conclusion: CTA was a useful tool and it was equivalent to MRA for the assessment of SST full thickness tear, SLAP lesion and PASTA lesion, but not bursal side partial tear. Thus, it may be used preferably to diagnose shoulder pathology and for follow up as an inexpensive tool after operation.