Seasonal variations in CH4 emission rates from rice paddies have been reported to have one or more maxima during the middle and late periods of rice growth. The factor affecting an appearance of CH4 emission maxima was examined in three types of pot experiments. In the experiment 1, four rice cultivars with difference in length of the period from transplanting to heading were transplanted on the same days. For the experiment 2, a cultivar was transplanted 4 times with interval of two weeks. In these experiments, the heading differed about a month between the earliest and latest treatments, respectively. However, shifting growth stage of rice plants did not shift the CH4 emission maxima, and the CH4 emission maxima often matched the maxima of daily mean air temperature. The effect of variation in temperature on CH4 emission rate was further investigated in the experiment 3 by placing the rice-planted pots under regulated temperature. Besides the first emission peak of CH4 attributable to rice straw (RS) carbon, three emission peaks corresponding to the peaks of air temperature were detected for the RS-applied pots placed outdoors. These three peaks were not observed or much less conspicuous for the RS-applied pots in a phytotron at 30¿C. Temporal decreases in CH4 emission were detected both for the pots placed in the phytotron and outdoors just after the topdressing of (NH4)2SO4, which was considered to be a major cause of irregular disagreement between the variations in CH4 emission rates and in air temperature during the middle period of rice growth. ¿ 2001 American Geophysical Union