OBJECTIVE To investigate the associations between large-joint damage and findings on fluorodeoxyglucose positron emission tomography combined with computed tomography (FDG-PET/CT) using the "assessment of rheumatoid arthritis by scoring of large-joint destruction and healing in radiographic imaging (ARASHI)" scoring system. METHODS A total of 270 large joints (shoulders, elbows, hips, knees, and ankles) in 27 rheumatoid arthritis patients were assessed. FDG-PET/CT was performed at the initiation of biologics. Radiographs at baseline and at 3 years were evaluated using the ARASHI score. RESULTS Radiographic progression of damage was detected in 35 by Larsen grade vs. 87 by the ARASHI score. The maximum standardized uptake value (SUVmax) at baseline, Steinbrocker stage at baseline, concomitant prednisolone use, and disease activity score in 28 joints based on erythrocyte sedimentation rate (DAS28-ESR) at 6 months were significantly higher in the radiographic progression group. An SUVmax higher than 1.65 at baseline was a significant predictive factor for progressive damage at 3 years. CONCLUSIONS The ARASHI score may allow more detailed evaluation of large joints than the Larsen method. Joint destruction is likely to have progressed at 3 years in large joints, which had a higher SUVmax at the initiation of biologics.