The difficulty in the determination of Free Core Nutation (FCN) parameters (period and quality factor) using high precision gravity technique is due to the low signal-to-noise ratio of observations, which affects the determination precision when using traditional least square method. In this paper, we use Bayesian method to fitting the FCN parameters with the help of gravity data observed by superconducting gravimeters in Wuhan and Global stations. The difference using observations from different stations is investigated by comparing the Bayesian method and least square method. We also discussed the influence on FCN determination of choosing different tidal wave and ocean tidal models. The results indicate that Bayesian method is more credible and the FCN quality factor obtained by Bayesian method has a better agreement with that from VLBI observations, which could offer effective reference for further studies of the deep interior of the Earth.