The complex relationship between the process parameters and the quality index of loosening and conditioning process was studied, and the network model was established by using Bayesian network analysis method. The results showed that: ① The network model of loosening and conditioning process could well reveal the influence relationship and influence weight of process parameters on the discharge moisture content and temperature. ② The influence degree of process parameters on the discharge moisture content was determined as from high to low, opening of automatic valve for air water mixing ＞ unit time material cumulative measurement ＞ water addition ratio ＞ opening of automatic valve for steam ＞ hot air temperature. But, the influence degree of process parameters on the discharge temperature was determined as from high to low, hot air temperature ＞ unit time material cumulative measurement ＞ opening of automatic valve for air water mixing ＞ opening of automatic valve for steam ＞ water addition ratio. ③ The prediction accuracy of network model of loosening and conditioning process on discharge moisture content and temperature were 64.34% and 65.72% respectively, which had a good application effect and practical value. It could be predicted that Bayesian network analysis method had a wide range of application prospects in guiding the actual production and improving the quality of cigarette processing.