Aiming at the random factors to equipment support at complex battlefield environment nowadays, introduce to use gray theory to solve the prediction of spare parts inventory during equipment support. Taking into account in dealing with small samples of traditional statistical data shortcomings, the use of Bayesian theory to predict the results of the evaluation, through the derivation of inverted gamma distribution of chi-square change, give a simple formula for the probability interval. Give a simple formula of probability intervals. Finally, an example is analyzed to prove that this method is better than statistical estimates, especially in the case of less historical data to predict effects obviously, which has practical guiding significance to equipment support.