The aim of this study was to compare the extent of in vitro T cell depletion and recovery of hematopoietic progenitor cells achieved with five methods of T cell depletion. Bone marrow samples from the same source were treated with monoclonal antibody Campath-1 (CP1) and human complement, XomaZyme-H65 (anti-T cell ricin A chain immunotoxin), or soybean agglutinin (SBA) alone or in combination with sheep erythrocytes (EAET) or a cocktail of immunomagnetic beads (B) directly coated with anti-CD2, anti-CD3, or anti-CD8 monoclonal antibodies. Residual T cells were enumerated by limiting dilution analysis, EAET rosetting, and proliferative responses to phytohemagglutinin. The results of this study demonstrated the following reductions in BM T cells as detected by limiting dilution analysis (mean % control): SBA+B (99.9%), SBA+EAET (99.8%), CP1+C' (99.4%), anti-T cell ricin A chain immunotoxin (99.0%), and SBA alone (94.2%). Neither PHA response nor enumeration of residual EAET rosettes provided discriminating differences in the degree of T cell depletion by treatment method when T cell reductions exceeded 99.0% by LDA. These results demonstrate the ability of CP1+C', XomaZyme-H65, and SBA plus sheep erythrocyte or magnetic bead depletion to achieve a greater than 99% reduction of BM T cells and the importance of limiting dilution analysis in defining differences in T cell numbers when depletion exceeded 99%.