PURPOSE: To evaluate the optimal pulse sequences in the detection of erosions and for the discrimination of theeffusion from pannus. MATERIALS AND METHODS: MR studies were performed in thirteen patients with clinical diagnosis of rheumatoid arthritis. We used a 1.5T MR unit with single or dual 3 inch coil. We compared coronal images at the same location, obtained with pulse sequences of T2, T2 with fat saturation, and postcontrast T1 with fat suppression. Erosions were assessed by comparing the count of the numbers of erosions depending on each imaging method. A four point scoring systems was used for the comparison of conspicuitof lesion. RESULTS: T2WI with fat suppression and postcontreat T1WI with fat suppresison were better than conventional T2WI in the detection of erosions. The pannus had a heterogeneous signal on T2WI, some of which were enhanced after the intravenous injection of MR contrast a gents. CONCLUSION: We suggest that the fat suppression TIWI with enhacnement or the fat suppression T2WI were better than conventional T2WI for detecting both bone erosion and pannus. The postcontrast MR imaging may often be helpul in distinguishing pannus from the effusion.