Abstract: To avoid the damage misidentification of traditional damage assessment methods induced by indeterminate factors such as measurement noise, a statistical damage assessment method is proposed. Regularization technique is introduced to alleviate the ill-posedness of the parameter estimation problem. A data perturbation method is employed to study the impact of measurement noise on identification results. The probability nature of estimated parameters is obtained by Monte Carlo method. The formula for statistical damage assessment is then established based on probabilistic method. The validity and accuracy of the proposed method were demonstrated through damage detection experiment and the damage detection results of both statistical and determinate assessment methods were compared. Experimental results show that the statistical damage assessment algorithm can improve damage misidentification notably in contrast with determinate method.