Abstract: Climate variability is one of the main factors affecting annual changes in crop yield. In this study, a risk analysis method was used to describe the relationship between climate change and crop yield. The results could provide the basis for reducing the risk of food production. Based on available data on yield of soybean and corn in 22 towns of Hailun City for 1978~2007, relative meteorological yield was used as an indicator for the influence degree of climate change on crop yield. Yield reduction probabilities for different crops in each township were calculated using information diffusion method. Furthermore, the risk of crop production affected by climate variability was identified. Using daily rainfall and temperature as aridity index, all years were divided into partially dry, normal and partially wet years. The degree of risk was estimated under different hydrothermal years for each township. Results show that the probability of extremely high-yield years is lower than that of extremely low-yield years. The risk level of production reduction of soybean in north area is lower than that in south area, while the corn shows the contrary spatial tendency. There are spatial differences in risk levels of crop production among the different hydrothermal years. And spatial distribution of soybean production is more relative with rive network than that of corn production.