The compound (125)IUdR can be incorporated in a stable form into the DNA of cells. The isotope is released if labelled cells or their progeny die. Consequently the rate of (125)I excretion from mice can be used to follow the fate of labelled cells in vivo. Using these principles we show:(1) Sufficient label can be incorporated in vitro into both fresh and cryopreserved human leukaemic myeloblasts, in non-toxic concentrations, to allow their survival in mice to be estimated by whole-body counting;(2) The release of isotope from labelled cells is sufficiently slow to offer reasonable expectation that this technique can be used for assessing the sensitivity of myeloblasts to cytotoxic agents in vivo (an application described in the second paper in this series, Sonis, Falcão and MacLennon, 1977);(3) The rate of (125)Iexcretion from mice injected with myeloblasts from different donors varies. This probably reflects different rates of spontaneous death of injected myeloblasts;(4) Active rejection of myeloblasts starts within 48 h of their injection into mice;(5) Indirect evidence that phagocytic cells may be active agents in myeloblast destruction in mice;(6) Various methods of immunologically depriving mice were assessed to see if they would result in a useful increase in survival of injected human myeloblasts. We conclude that there is little advantage and some limitations in using mice thus deprived;(7) One of the agents used for immunological deprivation-silica powder-markedly decreased the rate of (125)I loss from mice injected with labelled killed myeloblasts. This experience emphasizes the importance of including the killed-cell control in this assay.