Rheumatoid arthritis (RA) is known to be caused by autoimmune disorders and can be partially alleviated through Disease-Modifying Antirheumatic Drugs (DMARDs) therapy. However, due to significant variations in the physical environment and condition of each RA patient, the types and doses of DMARDs prescribed can differ greatly. Consequently, there is a need for a platform based on patient-derived cells to determine the effectiveness of specific DMARDs for individual patient. In this study, we established an RA three-dimensional (3D) spheroid that mimics the human body's 3D environment, enabling high-throughput assays by culturing patient-derived synovial cells on a macroscale-patterned polycaprolactone (PCL) scaffold. Fibroblast-like synoviocytes (FLSs) from patient and human umbilical vein endothelial cells (HUVECs) were co-cultured to simulate vascular delivery. Additionally, RA characteristics were identified at both the genetic and cytokine levels using real-time polymerase chain reaction (RT-qPCR) and dot blot assay. The similarities in junctions and adhesion were demonstrated in both actual RA patient tissues and 3D spheroids. The 3D RA spheroid was treated with representative DMARDs, observing changes in reactive oxygen species (ROS) levels, lactate dehydrogenase (LDH) levels, and inflammatory cytokine responses to confirm the varying cell reactions depending on the DMARDs used. This study underscores the significance of the 3D drug screening platform, which can be applied to diverse inflammatory disease treatments as a personalized drug screening system. We anticipate that this platform will become an indispensable tool for advancing and developing personalized DMARD treatment strategies.