This study deals with the interpretation of primary data of the 51Cr release assay for cellular cytotoxicity. In particular the dose-effect relationship between increasing numbers of lymphoid cells and the percentage of target cells killed has been considered. The number of target cells killed depends on the frequency and the activity of cytotoxic cells. These two parameters are often not distinguished from each other, which causes confusion and frequently results in vague descriptions of cytotoxicity data. Because in many cases not all cells in the target cell population can be lysed, we recommend the introduction of the plateau value for target cell kill. This maximum of target cell kill is a measure of the frequency of lysable target cells, but also depends on the cytotoxic cell activity. Description of the dose-effect curve by y = A(1 - e-kx) allows the simultaneous calculation of the maximum kill (A) and the frequency of cytotoxic cells (k) in the lymphoid cell population (x). Results are presented which indicate that A and k represent totally independent biological parameters the use of which permits a more objective description of cytotoxicity data.