A probabilistic damage size and location diagnosis framework is proposed in this paper. The proposed method integrates the Lambwave-based damage detection and a Bayesian updating method for damage detection and localization. First, finite element method (FEM) is used to simulate the lamb wave propagation within thin aluminum plate, in which the electrical potential response is collected by coupling the piezoelectric element with the mechanical element. Following this, an advanced signal feature interpreting technique is used to extract the damage features, such as the normalized amplitude change and correlation coefficient from the received signal. Next, Bayesian theorem is introduced and probabilistic damage size and location detection framework is developed. Posterior distributions of the damage size and location are obtained using Bayesian updating with identified damage features. Finally, the proposed methodology is demonstrated using for two numerical examples. Some conclusions and future work are drawn based on the proposed study.