Rheumatoid arthritis (RA) is related to alterations in different inflammatory and connective tissue biomarkers. The diagnostic values and the factors affecting these biomarkers are conflicting. In the present study, a bone-related composite (B-composite), made from the z-score of stromelysin-1 (MMP3), colony-stimulating factor 2 (CSF2), and osteopontin (OPN), and I-composite, reflecting immune activation, made from the z-score of tumor necrosis factor-α (TNFα), interferon-γ (INFγ), and vascular endothelial growth factor-A (VEGF) were examined in RA patients. The biomarkers were measured by ELISA technique in 102 RA patients and 58 age-matched healthy control subjects. Serum MMP3, TNFα, IFNγ, and CSF2 showed significant elevation in RA patients. Multivariate general linear model (GLM) analysis revealed a significant high effect of diagnosis on biomarkers' level (partial η2 = 0.415). Duration of disease is significantly associated with VEGF, OPN, and B-composite and negatively correlated with TNFα. B-composite is significantly associated with CRP. A significant fraction of the DAS28 score variance can be explained by the regression on zlnINFγ. The variance in the CRP was explained by zlnOPN and B-composite. More than half of anti-citrullinated protein antibodies (ACPA) variation can be explained by the regression on serum MMP3 and I-composite. The top 3 sensitive predictors for RA disease are INFγ, MMP3, and TNFα. B-composite is associated with the duration of disease and CRP. At the same time, I-composite is negatively associated with the ACPA level. The biomarker composites have potential use as RA disease characteristic biomarkers.