Background: It is common for osteochondral lesions of the talus (OLT) to be associated with ankle injury, residual ankle stiff ness and pain after osteochondral injuries, and OLTs are most common in the dome of the talus. Before choosing the appropriate therapy based on diagnosis, it is necessary to determine if the OLT is a stabile lesion or not. More accurate diagnosis and classifi cations could be conducted to guide the clinical treatment. Objectives: To compare the accuracy rate and diagnostic indices of proton-density fat-saturation BLADE (proprietary name for periodically rotated overlapping parallel lines with enhanced reconstruction in magnetic resonance systems from Siemens Healthcare) sequences from a small fi eld of view (FOV) surface coil and proton-density fat-saturation sequences without BLADE in a routine FOV using a boot-shaped coil on the magnetic resonance imaging (MRI) of OLT. Patients and Methods: One hundred and one patients with ankle pain for greater than 6 weeks and normal X-ray participated in this study. They were ready for diagnosis by arthroscopy and treatment following MR examination. All patients were scanned by two diff erent methods: (a) a coronal proton-density fat-saturated sequence with BLADE using a small FOV surface coil and (b) a coronal proton-density fat-saturated sequence without BLADE using a routine FOV boot coil. In this study, the arthroscopic results were regarded as the gold standard for assessing OLTs and were used to directly compare the ability of the two diff erent scanning methods to produce accurate fi ndings with regard to MR imaging of OLTs. Results: Of the 101 patients, 57 patients with cartilaginous injury were confi rmed by the results of arthroscopy, and 44 patients were without cartilaginous injury. Forty-eight lesions had the same grade with the small FOV BLADE sequence MRI and arthroscopy, while 37 lesions had the same grade on routine FOV MRI and arthroscopy among 57 patients with cartilaginous injury. Among MR fi ndings of 101 patients, the results of weighted Kappa, accuracy, sensitivity and specifi city were 0. 803, 90. 10%, 84. 21%, and 97. 73%, respectively between small FOV and arthroscopy; 0. 515, 75. 25%, 64. 91%, and 88. 64%, respectively between routine FOV and arthroscopy. The accuracy rate, sensitivity and specifi city of small FOV with BLADE were higher than routine FOV. Mc Nemar’ s test confi rmed that the accuracy, sensitivity and specifi city of small FOV BLADE were signifi cantly better than routine FOV (P < 0. 001). Greater values of small FOV compared to routine FOV appeared in sensitivity of 0. 94 vs. 0. 81 (MR normal and grade I), 0. 89 vs. 0. 77 (MR normal and grade I-IIA), 0. 90 vs. 0. 74 (MR normal and grade I-IIB), 0. 91 vs. 0. 75 (MR normal and grade I-III) and specifi city of 0. 98 vs. 0. 92 (MR normal and grade I), the same in specifi city of 0. 91 vs. 0. 91 (MR normal and grade I-IIA), 0. 80 vs. 0. 80 (MR normal and grade I-III), only smaller in specifi city of 0. 60 vs. 0. 80 (MR normal and grade I-IIB). Conclusion: Small FOV surface coil with BLADE sequence grading of OLT demonstrated a promising accuracy rate compared with arthroscopic diagnosis.