An acoustic resonance diagnostic method for detecting structural degradation is provided. The method includes steps as follows: Firstly, a training model is built using a deep neural network. At least two training acoustic signals are inputted to the training model to carry a training. A diagnostic model is built according to a result of the training using a CNN. A plurality of under-test sound wave signals are captured from an under-test section of an under-test structure. An AI is used to filter data invalid for detecting structural degradation from the under-test sound wave signals. A structural degradation state of the under-test section is determined according to a remaining portion of the under-test sound wave signals through the diagnostic model.