In order to solve the uncertain damage problem caused by measurement noise, a damage identification method based on acceleration inner product vector and gray cloud model is presented. First, basic theory and formulas of cloud model and cloud generators are introduced. Then, the acceleration response under random excitation load is calculated, inner product vector is deduced from cross correlation functions and second-order difference method, and an inner product vector damage index is proposed. Finally, the grey cloud rules of inner product vector and damage intervals are built. Considering the uncertainties casused by stochastic measurement noise, a damage identification method based on gray cloud model is presented by using weighted summation and averaging in various modes. Simulation results show that the identification results of the proposed method are better than those of the inner product vector damage index, and the damage identification method based on inner product vector and gray cloud can solve the uncertain damage problem caused by measurement noise.