The aim of this study was to assess the cytotoxicity of a new leucocyte-labelling method, which may be used clinically to localize inflammatory and immune reactions. Human blood leucocytes, their mononuclear sub-population, and mouse mononuclear bone marrow cells were labelled with 99mTc for 30-45 min, washed once, and then evaluated in various functional assays. The new procedure includes [99mTc]-labelling with a bisalt method, in the presence of dihydroxybenzoic acid as an intermediate antioxidant-complexing stabilizer, and a carboxylic acid salt of stannous ions as a reducing agent. To challenge the method, cells were labelled about two orders of magnitude more heavily in these initial methodological studies than in on-going clinical trials. Labelled leucocytes ingested latex beads as readily as the controls, but migrated chemotactically and randomly somewhat slower than the control cells. The lymphocytes were triggered by PHA and Con A in a normal way. However, lymphocytes and haemopoietic progenitor cells exposed to radiation for several days, were killed by the isotope doses used, of which about 2% (i.e. 20 MBq) were bound per million cells. All deleterious effects were apparently due to irradiation, and the labelling procedure itself did not damage the cells.