Graded numbers of bone marrow (BM) cells were injected into fatally irradiated mice. Eight days later the mice were given 3.0 microCi (1 Ci = 3.7 X 10(10) Bq) of 125IdUrd to label proliferating cells in the spleen and BM. On day 9 the mice were killed and the spleens and femurs were removed for splenic colony assay and measurement of radioactivity in the spleen and femurs. The number of splenic colonies shows a linear relationship with dose of marrow cells injected from 10(4) to 10(5) cells. The slope of the curve of spleen colonies versus number of cells injected is less than 1, implying that the fraction seeded in spleen decreases with number of cells injected. Above 10(5) and below 10(4) there is a striking departure from the simple linearity. Below 2 X 10(3) cells injected, the logarithm of the observed colony yield is linear with logarithm of the number of cells injected. Poisson calculation of the average number of pluripotent stem cells that should be present with numbers of marrow cells injected below 2 X 10(3) followed closely the actual observations. The data show that there is no detectible proliferation in the BM until the dose of marrow cells exceeds 3.5 X 10(4) cells. Induction of cells into cycle increases the seeding into the BM, and thymidine cytocide drastically reduces seeding in the BM, leading us to conclude that the BM is repopulated almost exclusively by stem cells in DNA synthesis.