In HLA-incompatible kidney transplantation, monitoring donor-specific antibodies (DSA) plays a crucial role in providing appropriate treatment and increases kidney survival times. This work aimed to determine if early post-transplant DSA dynamics inform graft outcome over and above other predictive factors. Eighty-eight cases were classified by unsupervised machine learning into five distinct DSA response groups: no response, fast modulation, slow modulation, rise to sustained and sustained. Fast modulation dynamics gave an 80% rate for early acute rejection, whereas the sustained group was associated with the lowest rejection rates (19%). In complete contrast, the five-year graft failure was lowest in the modulation groups (4-7%) and highest in the sustained groups (25-31%). Multivariable analysis showed that a higher pre-treatment DSA level, male gender and absence of early acute rejection were strongly associated with a sustained DSA response. The modulation group had excellent five-year outcomes despite higher rates of early rejection episodes. This work further develops an understanding of post-transplant DSA dynamics and their influence on graft survival following HLA-incompatible kidney transplantation.