Although human lymphocyte cultures contain cells that have divided for different numbers of times after PHA stimulation, this heterogeneity of different cell divisions can be explained by a difference in the times when cells start their first DNA synthesis in responding to PHA. Cycling lymphocytes, whether they entered cycling earlier or later after stimulation, have about the same mean cell cycle times of 12-14 hr. Treatment with 3 X 10(-6) M MMC was found to induce an approximately 5-hr delay in the cell cycle. The induction of SCEs by chemical treatment depends on the stage in the cell cycle at treatment and on the persistence of the induced SCE-forming lesions. The most efficient time of treatment is the G1/S boundary in the first cell cycle of the 2 consecutive cycles before sampling. Among 3 alkylating agents tested here, EMS and 4NQO induce quite long-lived lesions that lead to SCE formation, whereas MMC-induced lesions seem to be completely removed within a cell cycle. Treatments with increasing concentrations of the chemical induce a larger increase in SCE frequency and longer delays in the cell cycle. However, with a fixed experimental regimen, treatments with relatively higher doses cause a deformity of the dose-response relationship. The data also show that longer BrdUrd treatment before fixation results in a sampling of cells that have higher SCE frequencies. Repeated pretreatments of human lymphocytes with a very low concentration of MNNG render them resistant to a following challenge with MNNG, or ENU, but not with MMC. The data at present indicate that there is a quite large variability among blood donors in this induction of resistance in lymphocytes.