Medical strategy for rheumatoid arthritis (RA) has markedly advanced in recent years. The introductions of biologics and methotrexate as an anchor drug have made it possible to not only suppress pain and inflammation (clinical remission), but also to inhibit joint destruction (structural remission), leading to cure of the disease. In order to achieve this target, it is the most important to diagnose RA early and promote disease remission. However, since the condition and pathology are diverse among patients, optimal treatment for each patient is desired (personalized medicine). Treatment should be performed under consideration of the disease state such as activity, prognosis regarding joint destruction, and complications. It is also important to clarify the patient characteristics, such as responsiveness to the drugs and risk of adverse effects. Biomarkers, such as proteomics and pharmacogenomics (genetic polymorphism, etc.), are indispensable for personalized medicine. We have established a predictive model for methotrexate hepatotoxicity, consisting of 13 SNPs with a sensitivity of 100% and specificity of 89%, although the model should be validated with a larger-scale prospective study. RA is a multifactorial disorder with clinically heterogeneous features. Gene-environment interaction is closely involved in the production of anti-CCP antibodies (ACPA); thereafter, secondary stimuli of joints may lead to symptoms of RA. Joint injury, emotional stress, and infections often trigger the onset of RA. Cure can be achieved through complete remission by early aggressive treatment and returning to the pre-clinical state of RA with environmental improvement.