Purpose : To evaluate the relationship between superior labral dimension of the glenohumeral joint on direct MRA and presence of SLAP lesion. Materials and Methods: IRB approval was obtained and informed consent was waived for this retrospective study. Direct MRA studies of the shoulder in 296 patients (300 shoulders) with arthroscopic surgery were analyzed by two radiologists blinded to the arthroscopic results, which were used as gold standard. One of the radiologists reviewed the images twice (session 1 and 2) for the evaluation of intra-observer variability. Transverse and longitudinal dimensions of superior labrum on coronal T1-weighted images were measured as base and height of the inverted triangular-shaped superior labrum and compared between patients with SLAP lesions vs. non-SLAP patients. Presence of meniscoid labrum was noted. Statistical analysis was done using unpaired t-test. Results: Among 279 patients (283 shoulders), 122 patients (43.1%) had SLAP lesions. The mean base/height of superior labrum in SLAP and non-SLAP patients measured on T1-weighted MR image were 8.8 mm / 5.2 mm, 8.5 mm / 4.9 mm for reader 1; 8.2 mm / 4.9 mm, 8.1 mm / 4.5 mm for session 1 of reader 2; 8.0 mm / 4.8 mm, 7.6 mm / 4.3 mm for session 2 of reader 2. In SLAP group, the mean labral height was larger than non-SLAP group with statistically significant difference (p<0.05). Fifteen patients (5.3%) had meniscoid labrum according to operation records. Conclusion: In patients with SLAP lesion, the height of the superior glenoid labrum on oblique coronal image of MRA was slightly larger than non-SLAP patients. A larger height of superior glenoid labrum may be associated with SLAP lesions.