Both spontaneous and EMS-induced mutant frequencies were determined in cultured cells from V79 Chinese hamsters using three different experimental protocols. After optimal expression time was attained, mutation frequencies only remained constant when a protocol was used in whcih the cell density was maintained below critical values both before and during mutant selection. The identification of such a plateau allows, besides more reliable and reproducible estimates of mutation frequency, reduction in the size of experiments for quantitative evaluation of mutagenicity. Determination of mutation frequencies over a wide range of expression times becomes in fact unnecessary.