At the case where the amount of failure data is small and the acceleration model is difficult to be determined,it is difficult for the traditional model to make a more accurate assessment of the testing results.Based on the grey prediction theory,the constant-experience data of life-obeying Weibull distribution was analyzed.The stress-related weights were used to generate background values to complement the missing data,and an equally spaced gray prediction model was established to correct the parameters of the accelerated life prediction model in this work.The analysis of the example indicated that the gray acceleration testing data evaluation model had a small relative error and high prediction accuracy.