To elucidate the therapeutic effect of orthopedic surgical intervention (OSI) in difficult-to-treat rheumatoid arthritis (D2T RA) compared with non-D2T RA. A total of 534 recent surgeries were analyzed only in patients who had undergone OSI since 2016 and for whom a 12-month post-operative follow-up was available. D2T RA was determined according to the EULAR definition, and patients with D2T RA were matched to patients with non-D2T RA using propensity scores calculated by a logistic regression analysis. The Health Assessment Questionnaire-Disability Index (HAQ-DI), Disease Activity Index 28 (DAS28), face scale, and patient's assessment of general health (GH) were measured repeatedly at baseline and 6 and 12 months and were compared using a two-way analysis of variance. The HAQ-DI, DAS28, face scale, and GH showed significant post-operative improvements, and there were significant differences in the HAQ-DI and face scale scores between D2T RA and non-D2T RA. An additional analysis with DAS28 as a covariate showed no significant interaction for either, suggesting that these improvements in clinical assessment were due to OSI rather than improved disease activity. In the absence of an effective pharmacological treatment strategy, OSI may be an effective treatment modality for the management of D2T RA.