Objective. The costs of biologic treatment per patient with rheumatoid arthritis (RA) are approximately 100 times the costs of treatment with a combination of conventional disease-modifying antirheumatic drugs (DMARDs). Despite this, biologic agents have not been proven superior. We compared the effects of combination DMARD therapies with and without biologic agents as therapy for patients with RA. Methods. Eight randomized controlled trials published in 10 articles were selected from a systematic literature search of 1,674 identified studies and integrated in a meta-analysis. These trials compared combinations of DMARDs versus a tumor necrosis factor (TNF) inhibitor plus methotrexate. Two reviewers independently entered data into standardized extraction forms. The combined effect measures were compared by means of the inverse variance method (continuous data) and the Mantel-Haenszel method (dichotomous data) using a random-effects model. Results. The primary outcome, radiographic progression score, did not differ between the combination DMARD group and the TNF inhibitor group, neither during the second year (20.09 units [20.61, 0.44]) of treatment or during the first 2 years (0.66 units [20.12, 1.43]). There were significant differences in the radiographic progression score, the American College of Rheumatology criteria for 50% improvement (ACR50), and the ACR70 response criteria at 6 months in favor of TNF inhibitor treatment, but these differences were not present in patients treated with an initial steroid course and disappeared at 24 months, irrespective of the use of steroids. Conclusion. The difference between DMARD combination treatments, including or excluding TNF inhibitors, is small. Due to the enormous cost differences, RA guidelines should recommend combination DMARD treatment before initiation of TNF inhibitors.