Proliferation variables such as mitotic activity and the percentage of S-phase cells have been shown to be of prognostic value in many tumors, especially in breast cancer. However, some studies reported a decrease in mitotic activity caused by delay in fixation of the tissue. In contrast, other studies showed that the identifiability of mitotic figures decreases after fixation delay, but the total number of mitotic figures and also the percentage of S-phase cells remain unchanged. Most studies have been done on small numbers of experimental tumors, thus introducing the risk of selection bias. The aim of this study was to reinvestigate the influence of fixation delay on mitotic activity and cell cycle variables assessed by flow cytometry in an adequate number of resected human tissues to reach firmer conclusions. Resection specimens of 19 and 21 cases, respectively, for the mitotic activity estimate and the flow cytometric percentage of S-phase calculation were collected directly from the operating theater using lung, breast, and intestinal cancers and normal intestinal mucosa. The tissues were cut in pieces, and from each specimen, pieces were fixed in 4% buffered formaldehyde (for mitosis counting) as well as snap frozen (for flow cytometry) immediately after excision, as well as after a fixation delay of 1, 2, 4, 6, 8, 18, and 24 hours. Moreover, during the fixation delay, one series from each specimen was kept in the refrigerator and the second at room temperature. Thus, a total of 304 (19 X 16) and 336 (21 X 16) specimens were investigated for the mitotic activity estimate and the percentage of S-phase cells calculation, respectively. With regard to the estimation of the mitotic activity, both clear and doubtful mitotic figures were registered separately, obtaining an "uncorrected" and "corrected" (for doubtful mitotic figures) mitotic activity estimate. The percentage of S-phase cells was obtained by cell cycle analysis of flow cytometric DNA-histograms. The results showed that the quality of the material decreased during the fixation delay, as reflected by poorer cellular morphology in the hematoxylin-and-eosin-stained slides, resulting in more difficult identification of mitotic figures and a more time-consuming procedure with regard to the mitosis counts, but not in a worse intraobserver and interobserver reproducibility, which was acceptable. The reduction in quality of the tissues also was shown by the flow cytometric measurements because the coefficient of variation and percentage of debris increased after 4 hours or more of fixation delay. However, the mean values of the "uncorrected" mitotic activity and the "corrected" mitotic activity showed no decreasing trend; neither did the average percentage of S-phase cells. In conclusion, within the time investigated, fixation delay has no clear influence on the proliferation features studied. Because of the decreasing quality of the histological sections, resulting in more difficult identification of mitoses and interpretation of DNA histograms, fixation delay should be kept as short as possible, keeping the tissue at 4 degrees C until fixation.