Cell cycle analysis was performed on two established cell lines of human bladder transitional carcinoma, MGH-U1 and T24 cells. The cells were cultured on round cover slips for three days, autoradiograms were prepared after pulse-labeling with 3H-TdR and the fractions of labeled mitoses (FLM) were calculated every 2 hours through 36 hours. The FLM curves were analyzed by the empirical half-height level method as well as Mendelsohn-Takahashi's asymmetry method which is a global type of FLM analysis based on Takahashi's generalized model. The results did not show any distinct differences in the cell cycle parameters between MGH-U1 and T24 cell lines. A remarkable discrepancy, however, was observed between the cell cycle times analyzed by the asymmetry method and by the half-height level method. Although there is no evidence that the asymmetry method is biologically correct up to now, Mendelsohn and Takahashi tested this method repeatedly and mathematically proved that it had much less error than that of the traditional boundary method.