Three human tumor cell lines of widely differing radiosensitivity were used to examine the characteristics of the 3-[4,5-dimethyl(thiazol-2-yl)-3,5-diphery]tetradium bromide (MTT) assay and to select suitable conditions for its use in assessing the response of cells to ionizing radiation. The optimal concentration of MTT and the time of incubation of the cells with MTT were individualized for each cell line. The relationship between absorbance and cell number was not linear over the wide range of cell numbers that were used. A calibration curve of absorbance against cell number for each cell line was therefore used. Using the assay to quantify metabolically viable cells, growth curves of irradiated and unirradiated cells were constructed on days 0–14 after irradiation. Accurate surviving fractions could be calculated only when cells were in exponential growth. Using this modification to its interpretation, the MTT assay was able to provide a reproducible measure of survival, which compared well with clonogenic cell survival measurements. However, the necessity to optimize conditions of the MTT assay for each cell line severely limits its usefulness in determining the radiosensitivity of cells in primary human tumor cultures.