Graft-versus-host-disease (GVHD) remains the principal complication of allogeneic bone marrow transplantation. In animal models mature T lymphocytes have been shown to be responsible for GVHD and, therefore, in vitro treatment of donor bone marrow using monoclonal T cell specific antibodies and complement is currently being investigated as a strategy for the prevention of GVHD. In the present studies anti-T12 and anti-T11 monoclonal antibodies and rabbit complement were used to remove T lymphocytes from normal bone marrow. The efficacy of depletion was investigated by immunofluorescence assays and by in vitro culture of the residual cells using nonspecific mitogens or allogeneic B cells as the proliferative stimulus in the presence of lymphocyte-conditioned medium containing interleukin 2 (IL-2). Immunofluorescence analysis showed complete depletion of T12+ and T11+ cells after treatment with the respective antibodies and with the combination. Nevertheless, culture of treated bone marrow with phytohemagglutinin (PHA) or concanavalin A (Con A) and conditioned media containing IL-2 resulted in the proliferation of mature T cells (T3+, T4+ or T8+, T11+). Stimulation of treated marrow with allogeneic cells (Laz 388) resulted in the growth of a population with natural killer (NK) cell phenotype (T3-, T11+, NKH1+). The latter population was found to be strongly cytotoxic against K562 cells, a standard NK target. As expected, NK cells that are T11+ and T12- appeared to be more effected by in vitro treatment with anti-T11 than with anti-T12. A clonogenic assay was then used to quantitate the efficacy of target cell depletion in vitro. Three sequential incubations of bone marrow with either anti-T12 or anti-T11 plus complement resulted in depletion of 1-2 logs of clonogenic cells. Treatment with both antibodies concurrently resulted in elimination of 2-3 logs of clonogenic target cells. Although multiple treatments with both anti-T12 and anti-T11 were more effective than similar treatment with only one antibody, it remains to be established whether such combinations will be necessary in the clinical setting or whether more selective depletion of T cells without removal of NK cells might be optimal.