Abstract: The paper aims to provide meteorological forecast of development period of maize accurately and timely for local governments, agricultural sector and farmers. According to the temperature index of maize which sowed when daily average temperature through 10℃ steadily, and analysis of the interval number correlation of the average daily temperature through 10℃ and 5℃ steadily and through 5℃ steadily in Liaoyang city in 1956-2010, the regression equation was established by day series and interval number to predict the beginning day of the average daily temperature through 10℃ steadily. Based on the regression equation, sowing time was forecasted. By using the data of agricultural meteorological observatory in Dengta in 1981-2010, the relation of development speed of each period and weather conditions from sowing to maturity stage was analyzed. After inspection of the correlation for each period, the authors found that the correlation of alternate day number of development period and the accumulated temperature ≥ 10℃ was the best. Temperature susceptibility and photosensitive were strong from emergence to maturity, thereby the forecast model of sowing to flowering and jointing to maturation could be established to forecast flowering and maturation of maize. After inspection, it is found that the correlation coefficient, multiple correlation coefficient and F value of these forecast model all reach a significant level of 0.01, the variation trend of back substitution and actual value is consistent comparatively, the absolute error (ABSE) and the standard error (RMSE) range between 2.0 and 4.3 days, which can meet the needs of agricultural meteorological operation and services and have the popularization and application value.