The clonal hprt mutation assay in human Tcells is based on the assumption that wild-type cells and hypoxanthine-phosphoribosyl transferase (hprt) mutant cells survive and proliferate at the same rate during the expression phase which is required for the expression of in vitro-induced mutants. We have tested this assumption in a study of mutant frequency (MF) and proliferation rate at different time points during in vitro expansion of human Tcells in nonselective medium. Peripheral blood lymphocytes from 11 individuals were studied using standard cloning procedures to determine the cloning efficiency (CE) and the hprt MF by 6-thioguanine (TG) selection. Another cell portion from each individual was allowed to proliferate in bulk culture for 8 days in vitro, before measuring CE and MF as above. In the directly plated cell population the CE was 45% and the MF 18.7 ? 15.3 x 10-6 (mean 2 S.D.), whereas the in vitro expanded cell population showed a CE of 38% and a significantly reduced MF of 8.3 2 6.9 x 10" (P = 0.0033). Thus, the mean MF was 56% lower in the in vitro expanded than in the directly plated cell population. The experiment was repeated in another group of ten individuals with essentially the same result. In a third experiment, freshly prepared cells from two donors were allowed to grow for up to 15 days in bulk culture in vitro. Cell growth, CE, and M F were determined every third day. The MF decreased gradually, and at day 12-15 it was only 25% of the initial value. The total number of clonable cells increased 13-fold during the 15 days of in vitro expansion, while the mutant, TGresistant cell population increased only Sfold. These results suggest that human hprt mutant T-lymphocytes have a reduced proliferation rate compared to wild-type cells during in vitro proliferation. Thus, measurements of chemical and radiation induced MF with the Tcell clonal assay may underestimate the true MF by a factor of 2 or more.