Abstract: For a better understanding of the effects of light-temperature factors on the grain yield and yield components in different latitude areas, we have done six field experiments in Shanxi Hanzhong, Sichuan Wenjiang, Yunnan Mile of China and Hanoi Wenlin, Zhongbu Guang’ai, Saigon Long’an of Vietnam from 2005-2006. The results showed as follows: there was real linear regression between seven light-temperature factors and grain yield. The size order of standard regression coefficient was: cumulateive illumination intensity >duration of day>average temperature>average illumination intensity>maximum temperature>effective accumulated temperature >minimum temperature. There was extremely significant negative correlation between average temperature and spike length while there was significant positive correlation between average illumination intensity and spike length. The partial correlation coefficient between average temperature and number of effective spikes per plant was bigger (r = 0.4576). Except for duration of day, the correlation between other light-temperature factors and number of effective spikes per plant was significant. The partial correlation coefficient between duration of day and grains per panicle, seed setting rate were bigger which was 0.3746, -0.5599, respectively. The partial correlation coefficient between minimum temperature and 1000 grain weight was bigger (r = -0.1897) but the correlation between all the light-temperature factors and 1000 grain weight was not significant.