With mechanical system becoming more and more complex. It is very important to research time-dependent kinematic reliability. In order to improve the efficiency and accuracy of mechanism kinematic reliability analysis. By means of the analysis of mechanism kinematic error by probability method. Based on the theory of maximum entropy and random process,the reliability model of time-domain error is established. The object of study is a 6-axis robot mechanism. The static and dynamic kinematic error was synthesized. Kinematics simulation analysis was conducted by using ADAMS. The solution could be computed with Monte Carlo method. Select 100 error samples for the first 5 periods. The kinematic error was estimated by ARIMA model.The principle of maximum entropy was applied to derive probability density function. The time-dependent kinematic reliability was then estimated by the model of time domain error. Compared with the simulation method,The correctness and validity of the time-domain error reliability model in the study of kinematic reliability are verified,which can provide the reference for future research.