To design and improve the weather index insurance product for frost injury in apple florescence, Xunyi is selected as the representative apple planting county in the Weibei region of Shaanxi Province. The data of meteorology, apple flowering phenophase and yields from 1991 to 2020 are used for design and analysis. The meteorological yield of apples is separated by using moving average and grey prediction methods. The weather index based on extremely low temperatures and the frost injury process cumulative hazard during apple flowering is constructed for selection. The weather index for frost injury in apple florescence is determined by correlation analysis between weather index and yield reduction rate. The regression model between weather index and yield reduction rate is constructed. According to apple yield risk distribution characteristics, using the parameter distribution method, a variety of distribution models are selected to fit apple yield risk distribution. The optimal distribution model is selected by A-D test results and probability density function diagram. The pure premium rates are determined under different compensation trigger conditions. Results show that the correlation between cumulative hazard weather index in the frost injury process and yield reduction rate is higher than that between extremely lowtemperature weather index in representative counties and yield reduction rate. The yield reduction rate has a linear relationship with the ∑T_s weather index. Apple yield risk distribution fit to Cauchy distribution. The pure premium rates of weather index insurance products for frost injury in apple florescence are from 0.40 to 1.84 percent when yield reduction rates are 2,4,6 and 8 percent in Xunyi. The rationality and applicability of the weather index insurance product have been enhanced compared to traditional agricultural insurance products, which is beneficial to giving full advantages to weather index insurance.