In order to forecast fish pond’s water temperature and to guide local fisheries production in Honghu Area, water temperature and air temperature data of local typical fish pond in spring and summer of 2011 and 2012 were analyzed. The results showed that: with the deepening of the water level, the diurnal range of water temperature decreased. Generally speaking, the difference of surface and deep water levels’ temperature in spring was larger while in summer relatively smaller. The correlation analysis results of water temperature and air temperature showed that significant relationship between water temperature of the day and air temperature of the day and past 1 to 3 days. With the mean, maximum and minimum air temperatures of the day and past 1 to 3 days as predictors, the mean, maximum and minimum water temperature forecast models of spring and summer seasons were established separately by stepwise regression method, and the determination coefficient reached 0.84 to 0.98. Comparing the modeled water temperature with the water temperature data of May and July 2011, and in May the model fitted better. Established model all had high precision with average relative error value less than 4%. That was to say, the established water temperature models could be used to forecast the fish pond water temperature for Honghu Area, which could provide effective guidance and meteorological services for fisheries production activities.