The monoclonal interleukin-1beta (IL-1β) antibody canakinumab is approved for the treatment of systemic juvenile idiopathic arthritis (SJIA). Its efficacy has been proven in several trials, but not all patients show a complete and sustained response to therapy. We aimed to analyze the association of baseline serum biomarkers with treatment outcome in patients with SJIA treated with canakinumab.Serum samples from 54 patients with active SJIA without recent macrophage activation syndrome (MAS) treated with canakinumab in an open-label response characterization study were subjected to a multiplexed bead array assay. Interesting targets from these analyses were validated by ELISA. Clinical treatment outcomes included modified pediatric American College of Rheumatology (pACR) 30 and 90 responses, clinically inactive disease (CID) within 15 days of treatment, and sustained complete response, defined as pACR100 or CID within 15 days of treatment plus no future flare or MAS.In canakinumab naïve patients most biomarkers were elevated when compared with healthy controls at baseline and some rapidly decreased by day 15 (IL-1RA, IL-6, IL-18 and S100A12). Responders had higher IL-18 and IFN-γ levels and lower CXCL9 levels at baseline, emphasized by the IL-18: CXCL9 and IFN-γ:CXCL9 ratios. These ratios had significant accuracy in predicting treatment responses.Differential regulation of the IL-18/IFN-γ/CXCL9 axis is observed in patients with SJIA. Higher IL-18: CXCL9 and IFN-γ:CXCL9 ratios at baseline are associated with a better clinical response to canakinumab treatment in SJIA. Future studies are needed to validate these findings and determine their generalizability to patients with recent MAS.