AimsSARS‐CoV‐2 causes systemic immune dysfunction, leading to severe respiratory dysfunction and multiorgan dysfunction. Granulocyte and monocyte adsorptive apheresis (GMA) therapy is designed to regulate an excessive inflammatory response and has been proposed as a potential therapeutic strategy for coronavirus disease 2019 (COVID‐19). We aimed to investigate a targeted subset of granulocytes and monocytes to be removed after GMA therapy in patients with severe COVID‐19 infection.MethodsWe established an ex vivo experimental system to study the effects of GMA. Blood samples were collected into EDTA‐treated tubes and a mixture of blood samples and cellulose acetate beads was used in GMA. After GMA, blood samples were removed, and the granulocyte and monocyte subtypes before and after GMA were determined by CyTOF mass cytometry. To analyze mass cytometry data with a self‐organizing map, hierarchical clustering was used to determine the appropriate number of metaclusters from t‐distributed stochastic neighbor embedding.ResultsWe included seven patients with severe COVID‐19 and four age‐ and sex‐matched volunteers. Granulocyte subsets removed by GMA strongly expressed CD11b, CD16, and CD66b, and weakly expressed CD11c, consistent with mature and activated neutrophils. Monocyte subsets strongly expressed CD14, weakly expressed CD33 and CD45RO, and did not express CD16. These subsets were indicated to promote the release of inflammatory cytokines and activate T cells.ConclusionsThe identification of the granulocyte and monocyte subsets removed after GMA in patients with severe COVID‐19 may help explain the potential mechanism underlying the effectiveness of GMA in COVID‐19 and other inflammatory diseases.