Patients with rheumatoid arthritis (RA) are more susceptible to infections, which elevate the levels of relative cytokines. However, the ability of the cytokines levels to predicate bacterial infections in RA patients remains unclear. Here, we assessed the ability of the combination of serum cytokine levels and blood parameters to diagnose bacterial infections in RA patients. We measured the levels of a panel of serum cytokine and blood parameters in 168 RA patients and 81 healthy individuals. RA patients were divided into the bacterial infection (INFE) group (n = 76) and RA flare without INFE group (n = 92). Bacterial infection was confirmed by microbial culture, imaging, antibiotic response, and typical clinical symptoms. The discriminative ability of the combination of the cytokine levels and inflammatory parameters was assessed using the receiver-operating characteristic (ROC) curves analysis and a novel bioscore system. The levels of interleukin (IL)-6 (p = 0.006), IL-10 (p = 0.019), interferon (IFN)-γ (p = 0.033), CRP (p < 0.001), and ESR (p < 0.001) were higher in patients of the INFE group than in patients with RA flare, and the absolute numbers of CD19 B cells (p < 0.001) and CD4 T cells (p = 0.009) were lower. For discriminating bacterial infection, the combination of IL-6, IL-10, IFN-γ, ESR, CRP, CD19 B cells, and CD4 T cells, provided an area under the curve (AUC) of 0.827 [(95% confidence interval (CI): 0.760-0.881)], which was profoundly larger than that of IL-6, IL-10, IFN-γ, ESR, CRP, CD19 B cells, or CD4 T cells alone. In addition, we also developed a bioscore system based on the combination of these seven biomarkers. Seventeen (100%) patients with a bioscore of 0 were non-infected, while seven (100%) patients with a score of 7 had bacterial infections. The bioscore based on the combination of ESR, CRP, IL-6, IL-10, IFN-γ, CD19 B cells and CD4 T cells may be a promising and robust tool to diagnose bacterial infections in RA patients.Copyright © 2020 Elsevier Ltd. All rights reserved.