Oil-impregnated paper (OIP) bushing is essential equipment in power substations, and moisture intrusion is one of the major faults during its operation. In this article, a method is proposed to diagnose the moisture intrusion fault by the hidden Markov model (HMM) and the shuffled frog-leaping algorithm (SFLA). First, HMM is employed to characterize the probabilistic relationship between the moisture states and frequency-domain spectroscopy (FDS). Then, SFLA is adopted to optimize the locatable results of moisture intrusion point (MIP) in the bushing. Next, extensive simulated data are used for parameter training in the HMM-SFLA model library, with cross-validation for optimal parameters' configuration of HMM and SFLA. Finally, the proposed method was validated by experiment in laboratory and field measurements. The results show that the proposed method can effectively diagnose the MIP of OIP bushings.