Studies using chronic in vivo BrdU exposure, isolating primitive stem cells, and determining BrdU labeling, indicate that stem cells cycle. BrdU is also incorporated into DNA during damage/repair. DNA, which has incorporated BrdU due to cycle transit is heavier than normal, while the density of DNA with damage/repair incorporation is intermediate. DNA density of purified lineage-rhodamine low (rho(low)) Hoechst low (Ho(low)) stem cells or FDC-P1 cell line cells-was assessed in vitro, after exposure to cytokines and BrdU (cycling model) or cytokines and BrdU with bleomycin to induce strand breaks and hydroxyurea to halt cycle progression (damage/repair model). We determined DNA density using cesium chloride (CsCl) gradients and either fluorometry or dot blot chemiluminesence. DNA from BrdU labeled cycling Lin-rho(lo)Ho(lo) or FDC-P1 cells was heavier than normal DNA, while damage repair DNA had an intermediate density. We then assessed BrdU labeling of Lin-rho(lo)Ho(lo) cells in vivo. We found that 70.9% of lin-rho(lo)Ho(lo) cells labeled at 5 weeks. DNA density of these cells was low, in the damage/repair range, but similar results were obtained with stem cells, which had proliferated in vivo. Dilution of BrdU in in vitro culture of proliferating FDC-P1 cells also resulted in damage/repair density. We conclude that in vitro BrdU labeling models can distinguish between proliferation and damage/repair, but that we cannot obtain high enough in vivo levels to address this issue. All together, while we cannot absolutely exclude damage/repair as contributing to stem cell BrdU labeling, the data indicate that primitive bone marrow stem cells are probably a cycling population.