Abstract Background The phase I trial of the humanized anti-CD26 monoclonal antibody YS110 for CD26-expressing tumors was conducted recently. The present study identifies a potential prognostic biomarker for CD26-targeted therapy based on the phase I data. Methods Box and Whisker plot analysis, Scatter plot analysis, Peason product moment correlation/Spearman’s rank-difference correlation, Bar graph analysis, and Receiver Operating Characteristics (ROC) were used to examine the correlation between sCD26 titer variation with YS110 administration and tumor volume change, RECIST criteria evaluation and progression free survival (PFS). Mechanism for serum sCD26 titer variation was confirmed by in vitro experimentation. Results Serum sCD26/DPP4 titer was reduced following YS110 administration and gradually recovered until the next infusion. Serum sCD26/DPP4 titer before the next infusion was sustained at lower levels in Stable Disease (SD) cases compared to Progressive Disease cases. ROC analysis defined the cut-off level of serum sCD26/DPP4 titer variation at day 29 pre/post for the clinical outcome of SD as tumor response or PFS. In vitro experimentation confirmed that YS110 addition reduced sCD26 production from CD26-expressing tumor and non-tumor cells. Conclusions Our study indicates that serum sCD26/DPP4 titer variation in the early phase of YS110 treatment is a predictive biomarker for evaluating therapeutic efficacy.