The unfalsified control is a type of data-driven, model free control method. Based on input-output data online computing and learning controlled variable to match current system states and acting on system, the dynamic and static quality of system requiring can be acquired which is used to test whether it satisfies performance index. Furthermore, based on unfalsified control theory, the problem of trajectory tracking control of mobile robot is analyzed. Finally, by adopting unfalsified control strategy to nonholonomic dynamic system directly, tracking desired trajectory is achieved rapidly and accurately.