Brain-computer interface(BCI) was a communication system that did not depend on the brain's normal output pathways of the peripheral nerves and muscles.BCI provided a new communication channel for the patients with severe neuromuscular disorders,and also had potential values in virtual reality,entertainment and aviation.Recently,BCI developed rapidly,and its researches mostly concentrated on feature extraction and translation algorithm.Cortical evoked potentials such as steady-state visual evoked potential(SSVEP) and P300 were widely applied in BCI because of simple feature extraction and requiring little training.The progress in SSVEP-based and P300-based BCI was reviewed,and its experimental paradigms and limitations were analyzed in this paper.