Abstract: The corrosion and fatigue processes of high-strength steel wire exhibit strong randomness, making it difficult to accurately predict their corrosion fatigue life. To quantify the uncertainty of parameters in the empirical model of high-strength steel wire corrosion fatigue life, two corrosion fatigue life prediction methods using classical and hierarchical Bayesian inference are proposed in this study. Firstly, the classical and hierarchical Bayesian methods are adopted to infer the parameters in the empirical model of the corrosion fatigue life of high-strength steel wire, quantifying the uncertainty of parameters, and obtaining the posterior distribution of parameters. Then, using the posterior distribution of parameters, the probabilistic corrosion fatigue life (p-C-S-N) surfaces under different survival probability are obtained by propagating the uncertainty of parameters, achieving the prediction of the corrosion fatigue life of high-strength steel wire. Finally, experimental data was employed to verify the effectiveness of the two prediction methods, and a comparative analysis of the two prediction results was conducted. The research results show that both classical and hierarchical Bayesian inference can accurately predict the corrosion fatigue life of high-strength steel wire. When the survival probability is p=50%, the prediction results of the two methods are basically consistent. When the survival probabilities are p=2.5% and 97.5%, the prediction interval of hierarchical Bayesian is greater than that of classical Bayesian, and the corrosion fatigue life predicted by hierarchical Bayesian is more conservative. The hierarchical Bayesian method is more general, and classical Bayesian method is a special case of hierarchical Bayesian method.