The distribution of radioisotopes in tissues was measured following i.v. injection of labelled thoracic duct lymphocytes into syngeneic rats. The rate of elution of an isotope from the labelled cells and the subsequent fate of the eluted isotope were shown to be the most important factors limiting the usefulness of such isotopes for measuring cell localization particularly in non-lymphoid tissues. Comparison of labelling procedures using [3H] and [14C]uridine, [3H] and [14C]leucine, [75Se]-L-selenomethionine, [99mTc]sodium pertechnetate and [51Cr]sodium chromate in vitro and [3H]thymidine in vivo showed that 51Cr had the fewest disadvantages in the present context. Using 51Cr-labelled cells, the radioactivity was measured in a wide range of non-lymphoid tissues, and estimates of cell traffic were obtained. In skin, for example, the results indicate a cell flux in the range of 10(4)-10(5) lymphocytes/gm/hr. Evidence is presented which suggests that the early substantial localization of labelled cells in the lung is not an artefact due to sequestration or embolization of traumatized cells but probably reflects a slow intravascular transit time through this capillary bed. The primary lymphoid organs, thymus and bone marrow were shown to include a subpopulation of lymphocytes which belong to the recirculating pool. The thymus always contained a greater concentration of radioactivity at 24 hr than all non-lymphoid tissues except liver and kidney (approx. 0-1% of the recirculating lymphocyte pool) and the bone marrow was capable of temporarily accepting a substantial proportion (approx.25%) of the injected cells.