Abstract: In order to monitor and warn the disaster of low temperature in autumn, and assess the impact of the disaster on unfilled rice rate, the author conducted the study. Using many years observation data of rice in Guizhou Province, the author analyzed the correlation of unfilled rice rate and meteorologic elements by statistic methods. The results showed that: the average temperature in heading and flowering period impacted unfilled rice rate seriously, and temperature in different stages had different influence on rice. Rainy weather caused rice rate decrease. Negative accumulated temperature was better than usual index to represent intensity of low temperature in autumn. Meanwhile, in order to provide basis for monitoring and early warning of low temperature in autumn in Guizhou, meteorological model of rice yield loss also was established in this paper.