In the traditional optimization design of sliding bearing,the influence of uncertain factors on sliding bearing was ignored.A novel reliability optimization design method for sliding bearing was presented.Due to the cognitive uncertainty of some parameters,probability theory and evidence theory were used to represent the distribution of uncertain parameters.By combining the probability theory,evidence theory and firstorder secondmoment method,the reliability optimization design model of sliding bearing based on hybrid theory was established.The genetic algorithm was used to optimize the design of sliding bearing.The results of an example show that the capacity of sliding bearing is increased by 60%,the friction coefficient is reduced by 10%,and the calorific value is reduced by 6.9%.Although the optimized design parameters based on reliability optimization design method are slightly reduced compared with the deterministic optimization design method,but the lubrication reliability of sliding bearing is significantly improved,which can reduce bearing fault caused by lubrication failure and has high engineering practicality.