Abstract: Abstract: This study on runoff forecast and irrigation water use has important significance in optimal allocation and rational utilization of water resources and the development of regional social economy programming. A lot of research work about runoff forecast and irrigation water use has been done by many famous experts, however they usually ignored the uncertainty in the process of research, and rarely forecasted the future irrigation water use by linking the determination of irrigation water use and runoff forecast. In this paper, the grey-time series analysis method was applied to forecast the annual runoff in CaiQi. This is a relatively new model, and it is a medium and long-term forecasting mathematical method of the traditional time series decomposition calculation combined with grey system theory. When calculated, the annual runoff time series is decomposed into 3 parts, trend item, periodic item and random item. Extrapolating to the future time according to the variation of each part, and then the forecast value is the sum of 3 parts. Among them, the trend item is calculated by the grey forecast method. Then based on above results, the predictions of the maximum economic benefit and the main crops' irrigation water at Hongyashan irrigation region under uncertainty in 2011-2020 were presented by solving the optimal allocation of water resources model under meteorological factor uncertainty in the irrigation area. The results indicated that the runoff forecast results were close to the measured values, and the eligible rate was 67.6%. This shows the applicability of the proposed method, but we also see that the prediction accuracy is not very high. This is because human activities have much influence on the runoff at Caiqi. It resulted in a large fluctuation of annual runoff. Moreover, in order to curb the reduction of runoff, a number of improvement measures have been adopted in recent years. So the runoff had a tendency to rebound after 2000 at Caiqi. These measures affected the accuracy of the prediction method. Moreover, effective measures for water-saving were necessary in arid areas. It could greatly improve the water use efficiency, and guarantee economic benefits. Compared to spring wheat, seed melon and honey dew melon, cotton had the biggest economic benefits by unit of water in the Hongyashan irrigation region, and the second was honey dew melon. So when the available water is lacking, in order to reduce the economic losses caused by lack of water for irrigation, the irrigation water of cotton and honey dew melon should be ensured first. This study can promote the practical application of the uncertain methods, reflect the complexity and uncertainty of the actual situation, and provide more reliable scientific bases for using water resources economically, fully improving irrigation efficiency, and keeping up the sustainable development of the irrigated area.