Normally, research concerning disaster risk evaluation is based on the occurrence probability of one or more disasters. There is little direct association between most disaster risk evaluation indexes and crop production. It is a common phenomenon for disaster to occur with little harm to crops, and this makes it difficult to evaluate disaster risk. Most research on the risk of production variation is simply from the perspective of production variables. Rarely is research conducted that evaluates the risk of different degrees of production reduction. Using the risk of rice yield reduction in Liaoning Province as an example, we adopt a linear sliding average method to calculate meteorological yield. Through the discrimination of normal distribution and the normalization of skewed distributions, this method provides the reduction rate in lean years, a variation coefficient in disaster years, and the spatial distribution law of risk probability in Liaoning Province at 5% and 10% reduction rate. Integrating risk evaluation indexes, we also processed the regionalization of rice yield disaster risk in Liaoning Province using K-Means, which divided Liaoning Province into low, moderate, less high, and high risk areas. The results showed that the reduction rate of rice yield per unit in lean years was between 4.6% and 17.51%. Overall, this was lower in the central and eastern areas and higher in the northeast and southwest. The variation coefficient for rice yield per unit in disaster years was between 0.251 and 0.965. There was a zonal distribution along the northwest to southeast direction, being least in the central and eastern areas, and increasing in the southwest and northeast. The regionalization of rice yield disaster risk was similar when the reduction rate was greater than 5% or 10%. One of the probability ranges was between 22.2% and 42.7% and the other was between 5.4% and 33.3%. These values were low in the central area, and higher in other areas. The different risk areas were scattered overall and only contiguous in a small area. There were 13 areas with high risk, which were distributed in the west and northeast of Liaoning Province. The number of areas with less high risk was 15, and these areas were mainly in south central Liaoning Province. There were 15 and four moderate and low risk areas, respectively, which were mainly distributed in central and southeastern Liaoning Province. Finally, we analyzed the relationships between rice yield reduction and climatic characteristics in all regions in Liaoning Province, and discussed solutions in terms of disaster prevention and reduction. For drought disasters, we could adopt methods to reduce losses, such as loosing fill in the surface soil, promoting water saving irrigation, applying mulch and straw covering techniques, using drought inhibitors or water retaining agents, increasing income and reducing expenditures, impounding in advance, and choosing drought-tolerant species. For rice chilling damage, possible solutions are choosing cold-tolerant species, using warmth-retaining measures such as mulching and smudging, irrigating in advance, and applying deep-water irrigation.