Mouse spleen cells sequentially agglutinated by soybean agglutinin (SBA) and peanut agglutinin (PNA) previously had been shown to be sufficiently depleted of graft-versus-host activity to allow reconstitution of lethally irradiated allogeneic recipient mice. We have now tested the extent of T-cell depletion in this cell fraction by various in vitro assays, including cytotoxicity testing with anti-Thy-1 antiserum, mitogenic response to the T-cell mitogens concanavalin A and phytohemagglutinin and allogeneic responsiveness in the mixed lymphocytes culture assay. By these criteria the SBA+, PNA+ spleen fraction, used previously in the in vivo experiments, was found to possess about 1% T-cell contamination. The slight contamination with T cells previously found in the singly agglutinated SBA fraction can be removed by a second fractionation with SBA, thus eliminating the possibility that a minor T-cell subpopulation bears receptors for SBA. Finally, we demonstrated that the twice-agglutinated fraction, by SBA and PNA or by SBA alone, contains a significant number of prothymocytes, thereby indicating that mouse prothymocytes bear receptors for both SBA and PNA. The implication of these findings to bone marrow transplantation in humans is discussed.