Abstract: In this study, a total of 73 groups of paddy soil and rice samples were collected from 3 counties of Xiangtan City, Hunan Province, which were Xiangtan, Yuhu, and Xiangxiang, respectively. The total content of Cd in soil and rice samples was determined, the available Cd in soil samples also was measured, and the correlation analysis was conducted among them, respectively. According to the prior probability and data feature, we proposed to use the Bayesian method to eva-luate the risk of Cd content exceeding criterion in rice and construct the risk assessment model. The results indicated that, the average soil Cd content, soil available Cd content and the rice Cd content in the study area were (0.891±0.638) mg·kg^-1, (0.791±0.582) mg·kg^-1 and (0.376±0.395) mg·kg^-1, respectively, and the exceeding standard rate of Cd content in soil and rice reached 89.0% and 52.1%. The rice Cd content was significantly correlated with the available Cd content and total Cd content in soil in Xiangtan and Yuhu. A better correlation was presented between the soil available Cd content and rice Cd content. The rice Cd content in Xiangxiang had no correlation with total Cd content and available Cd content in soil. When the total Cd content and available Cd content were selected as the variables of Bayesian model, the deviations between the predicted exceeding probability of Cd in rice and the true exceeding standard rate were 6.6% and 3.9%, respectively, which proved the Bayesian risk assessment model was reliable. Selecting the available Cd content as the model variable, the predicted result was closer to the true value.