Existing research on college students' psychological problems mainly stays at the mental health education level.The analysis methods and corresponding treatment measures of psychological problems have been sometimes proposed.However, it is still difficult to accurately predict the students' psychological crisis, and is lacking in effective means to discover these problems.To settle this, a hidden Markov model (HMM) was applied to predict the psychological crisis of college students.A core factor set which may lead to crisis was constructed based on the theory of vague set, and the factors that affect the mental health of college students were analyzed.An initial HMM to predict college students psychological crisis was built by figuring out the initial value of each element of the model, and trained by using the Baum-Welch algorithm.Combining with the Viterbi algorithm, the prediction was carried out, and its correctness was verified by instance analysis according to the original students' psychological health valuation records in an university.The results show that the method can improve the prediction accuracy by adopting the core factor set in the model which would be an effective means to foresee the hidden crisis.