This methodologic study was performed on a single-cell-cycle breast carcinoma to evaluate the feasibility of computer-aided S-phase fraction determination in DNA static cytometry. The investigation was performed on Feulgen-stained cytologic material in which the total optical density values of 1,000 consecutive, randomly selected nuclei were analyzed (MultiCycle software). A good correlation in the S-phase fraction value with flow cytometry was obtained when the G2/G1 ratio was fixed at 1.95, when the histogram data points were smoothed at least once and the coefficient of variation of the G2 peak was the same as that of G0-G1 or when a first-order S-phase polynomial model was used. The percentages of nuclei in G0-G1 and G2 were somewhat similar to those obtained with flow cytometry. The greatest discrepancy with flow cytometry was observed in the value of the coefficient of variation of the G0-G1 peak of the static cytometric data: it was at least twice as great. It always remained high despite the software options used. As for the influence of the sample size in the S-phase calculation, the software was also run on samples of 600 and 200 nuclei. When the G2/G1 ratio was fixed at 1.95, the data obtained from 600 nuclei did not differ from those obtained with 1,000 nuclei, whereas an analysis on 200 nuclei showed a substantial variation. The software also allowed calculation of the ratio of the G0-G1 peak of the neoplastic population against that of the diploid reference (DNA index), the value of which in flow cytometry was 1.0.(ABSTRACT TRUNCATED AT 250 WORDS)