DNA ploidy and cell-cycle distribution were determined by flow cytometry in fresh tumour tissue of 53 cervical carcinomas. Epithelial cells were labelled by a fluorescein-isothiocyanate-conjugated cytokeratin antibody (CK6, CK18) to study the influence of contaminating stromal and inflammatory cells on results of cell-cycle analysis of tumour cells. Without identification of cytokeratin-positive cells 30/53 (57%) tumours were found to be DNA-aneuploid compared to 43/53 (81%) after gating for cytokeratin. Only 7 of 15 DNA-multiploid tumours could be detected without cytokeratin staining. In addition, cytokeratin-negative cells, which are found in all tumours, can be used as an internal standard for the calculation of ploidy and for quality control (coefficient of variation, linearity) of each individual sample. Cell-cycle analysis revealed significantly higher S-phase and G2M-phase fractions in cytokeratin-gated compared to ungated samples (13.1% versus 10.0% and 8.0% versus 5.4%; P < 0.001). This difference was more pronounced in DNA-diploid than DNA-aneuploid tumours. In conclusion, about 30% of DNA-aneuploid tumours could only be detected after cytokeratin labelling of epithelial cells. Owing to the identification of cytokeratin-positive cells the influence of non-tumoural cell elements on cell-cycle analysis was reduced markedly. Therefore, in cervical cancer, cytokeratin labelling can optimize both the determination of DNA ploidy and cell-cycle analysis.