Background: In organ transplantation, apheresis is frequently used for removal of anti-HLA antibodies. However, it is unclear whether plasmapheresis (PP) or semi-selective immunoadsorption (IA) should be employed, and the optimal number of apheresis sessions required to reach post-treatment objectives is also unknown. Methods: We enrolled 43 patients from Bordeaux University Hospital who were treated with PP (n=29) or IA (n=14) for antibody-mediated rejection or pre-transplant desensitization. Using Luminex single-antigen flow beads, we assessed the initial mean fluorescence intensity (MFI) of 1416 positive beads with MFIs obtained after 7 to 8 apheresis sessions (extended protocol) and, if a serum was available, after the first four sessions (short protocol). Results: MFI reduction after extended apheresis protocol was stronger with IA [87% (61%-100%)] than with PP [73% (22%-100%)] (P <.001). Indeed, 59% of the beads had a final MFI < 2000 with IA, whereas only 38% with PP (P <.001). The efficacy of removal depended on initial MFI but not on HLA specificity. A short protocol of apheresis showed excellent results without superiority of IA over PP for antibodies with an initial MFI < 3000. For antibodies showing MFI ?2000 after four sessions, the residual MFI predicted the effectiveness of four additional sessions. Conclusion: Monitoring the MFI of anti-HLA antibodies before and during apheresis protocol can guide physicians in the selection of apheresis technique and the number of sessions to be performed.< Réduire