The present invention relates to the field of natural human-computer interaction technologies, and discloses a bioelectricity-based control method and apparatus, and a bioelectricity-based controller, so as to improve naturalness of human-computer interaction. The method is as follows: performing characteristic extraction on a collected surface electromyography signal generated when a user performs a finger press operation, so as to obtain characteristic information; determining, according to a pre-created finger type recognition template, a finger type that is used to perform the finger press operation and that is corresponding to the obtained characteristic information; and mapping the determined finger type used to perform the finger press operation to a corresponding first instruction, and controlling a controlled device according to the first instruction. In this way, the controlled device may be controlled in a more harmonious and natural human-computer interaction manner.