The study was purposed to investigate the effect of bone marrow graft collection on the composition of peripheral blood stem cell grafts in healthy donors after recombinant human granulocyte colony-stimulating factor (rhG-CSF) application in vivo. Sixty-two healthy donors were treated with rhG-CSF [5 microg/(kg.d)] injected subcutaneously for five consecutive days. Donors were divided into groups A and B, and 31 donors were there in each group. Bone marrow grafts and peripheral blood stem cell grafts were harvested on the day 4 and 5 respectively in group A. In group B, only peripheral blood grafts were collected on both the day 4 and 5. The quantities of the cell components, CD3(+), CD3(+)CD4(+), CD3(+)CD8(+), CD14(+), CD34(+) cells and CD3(+)CD4(-)CD8(-) T cells were determined by multi-color flow cytometry. The results showed that median counts of nuclear cells (1.56 x 10(5)), lymphocytes (8.56 x 10(4)), CD3(+) (6.12 x 10(4)), CD3(+)CD4(+) (3.38 x 10(4)), CD3(+)CD8(+) (2.27 x 10(4)), CD14(+) (3.83 x 10(4)), CD34(+) cells (744) and CD3(+)CD4(-)CD8(-) T cells (3588) per microliter of peripheral blood stem cell grafts in group A were similar to counts of nuclear cells (1.40 x 10(4)), lymphocytes (7.34 x 10(4)), CD3(+) (5.32 x 10(4)), CD3(+)CD4(+) (3.06 x 10(+)), CD3(+)CD8(+) (1.83 x 10(4)), CD14(+) (3.21 x 10(4)), CD34(+) cells (554) and CD3(+)CD4(-)CD8(-) T cells (3120) in group B (p > 0.05). There were no difference in the ratios of CD4(+) cells/CD8(+) cells, CD14(+) cells/CD3(+) cells and CD3(+)CD4(-)CD8(-) T cells/CD3(+) cells in peripheral blood stem cell grafts between group A [1.52 (0.54 - 2.87)], [0.57 (0.15 - 1.64)], [0.064 (0.018 - 0.673)] and group B [1.68 (0.31 - 3.35)], [0.59 (0.18 - 1.25)], [0.063 (0.021 - 0.136)] (p > 0.05). It is concluded that no effect of bone marrow graft collection on the composition of peripheral blood stem cell grafts in the same donor is found after rhG-CSF application in vivo, bone marrow grafts and peripheral blood stem cell grafts can be collected respectively or simultaneously.