Provided herein are methods for predicting a response to anti-TNFα biologic treatment in patients suffering from rheumatoid arthritis (RA). The methods can be used to discriminate or categorize patients as likely to respond to a monoclonal antibody-based anti-TNFα biological inhibitor, but not likely to respond to an antibody fragment-based anti-TNFα biological inhibitor or a TNF receptor-based biological inhibitor. Alternatively, the methods are also useful to determine whether patients with RA are likely to respond to an antibody fragment-based anti-TNFα biological inhibitor or a TNF receptor-based biological inhibitor, but not likely to respond to a monoclonal antibody-based anti-TNFα biological inhibitor. Also provided are methods for treating a patient with RA based on the patients level of rheumatoid factor (RF) and/or anti-cyclic citrullinated peptide autoantibody (ACPA) relative to reference control levels.
