Fab arm exchange by a stabilized anti-IL-31 IgG(4)S228P monoclonal antibody (mAb) was studied using physiologically relevant antibody concentrations and thiol exchange conditions, and directly compared to that of matched wild type IgG(4) (IgG(4)wt) and IgG(1) control antibodies. In vitro arm exchange between the test mAbs and a purified IgG(4)wt exchange partner was monitored using capillary isoelectric focusing and a size-exclusion peak shift assay. Arm exchange between the test mAbs and IgG exchange partners with unknown specificity was monitored using only the shift assay. Studies were performed using single isotype human and mouse mAbs, unfractionated human, mouse, and cynomolgus monkey IgG, and human serum as the sources of the exchange partners. In vitro studies using human serum demonstrated that anti-IL-31 IgG(4)S228P did not undergo significant Fab arm exchange with endogenous human IgG(4) whereas anti-IL-31 IgG(4)wt underwent rapid and extensive Fab arm exchange. The in vitro results were corroborated by in vivo studies in which mice were injected with a mixture of either form of the test mAb and an excess of non-specific human IgG(4) exchange partner.