An in vivo study of the hamster cheek pouch epithelium using the stathmokinetic technique (Colcemid®) demonstrated a circadian variation in mitotic influx. Based on measurements of all nucleated epithelial cells the diurnal mean was estimated in two separate experiments as 0.34%/h ±0.02 (SE) and 0.27%/h ±0.02 (SE) respectively.³HTdR was injected in the latter study (a double labelling experiment). The significant difference between the two experiments is, however, probably due to biological variations. The maximal values for the mitotic rate were found during the light (resting) period, as were the maximal values for the mitotic index. The mean mitotic influx for the 'light period' was estimated as 0.5-0.4%/h, and for the 'dark period' as 0.2%/h. Independent analyses demonstrated the necessity of a circadian-dependent correction of the 1 and 4 h values of accumulated metaphases. The 1 h value was significantly too high during the light as well as the dark period. The 4 h value was found to be too low, but only significantly so during the dark period. Basing the estimation of mitotic rate on the 3 h accumulation value produced only very similar results to those found by using all four accumulation periods. The use of overlapping experiments proved that only cells entering mitosis after Colcemid application were arrested, so that when arrested metaphases were counted the accumulation line was correctly drawn through the origin. In the latter study (the double labelling experiment) both S-(Møler and Keiding 1982) and mitotic influx were estimated, the estimates being 0.55%/h ±0.03 (SE) and 0.27%/h ±0.02 (SE) respectively. Even considering possible methodological problems, the discrepancy between the S efflux and the mitotic influx indicates cell death and/or differentiation from G₂.