Using software programs provided by Coulter Electronics, we have developed an analysis system that would address problems encountered in DNA flow cytometric analysis of heterogeneous solid tumor populations, especially where the G2-M phase of the diploid population contaminates the S-phase of the aneuploid population, causing an overestimation of cells in S phase. We used the PARA 1 and PARA 2 programs in concert and developed three analysis models: (a) for euploid tumors; (b) for hyperdiploid tumors with overlapping populations; and (c) for near-diploid aneuploid tumors. Our purpose in this paper is to determine the limits and reproducibility of this analysis system with an emphasis on tumors with overlapping populations. Aliquots of frozen, pulverized breast tumor tissue (50 to 100 mg), routinely used in the steroid receptor assay, were used for routine flow cytometric measurement of the DNA index and S-phase fraction. To determine the accuracy of the analysis when overlapping populations were present, we mixed an aneuploid breast cancer cell line with human blood lymphocytes in varying ratios. A 10% mixture of aneuploid cells, the lowest mixture tested, still allowed analysis results within 95% confidence limits. Reproducibility of the system was assessed on frozen breast tumor tissue by intra- and interassay variation studies measuring cell cycle parameters and coefficient of variation of the G0-G1 peak width. Within any sample the amount of variation (+/- 2 SD) for the G0-G1 value was +/- 4.40 for intraassay and +/- 4.60 for interassay, and the amount of variation for S phase was +/- 3.0 and +/- 3.2 for intraassay and interassay, respectively. There was no difference in the variation of estimates for G2-M (+/- 2.6 for both intra- and interassay). In this study, the coefficient of variation of the G0-G1 peak greater than 5% was defined as unacceptable for accurate analysis, with the conclusion that S-phase fractions in aneuploid tumors can be routinely analyzed in human breast tumor biopsies despite tumor cell heterogeneity.