A number of assays exist for hematopoietic stem cells in both humans and mice, but the appropriate stem cell assay for the repopulating potential of human marrow is not clear. Two murine models suggest that these assays may not always predict marrow proliferative potential. In vivo diffusion chamber culture growth of CD1 marrow depleted of pluripotent stem cells (CFU-S) by exposure to mouse-brain antisera plus complement was equivalent to or greater than that of normal serum treated control marrow. Furthermore, CF1 mice repeatedly injected with endotoxin had markedly stimulated granulopoiesis with increases in the number of marrow CFU-S and the % in S phase but no changes in the number or proliferative status of marrow CFU-C. However, inbred BDF1 mice chronically injected with endotoxin although also showing striking increases in granulopoiesis had no significant alteration in their marrow CFU-S or CFU-C number or cell cycle status relative to saline injected controls. Both models present examples where conventional stem cell assays do not provide insight into marrow cell production and suggest that in vitro clonal assays of human marrow cells may not always predict for the potential of marrow to repopulate a human transplant recipient.