Abstract: To overcome shortcomings of poor stability and maneuverability of general bionic underwater robots, a bionic circular-fin propeller and its control method are proposed. A method of motion control based on central pattern generator (CPG) is proposed according to the structure features and propulsion mechanism of circular long-fin undulating propeller. A modeling concerning about 20 frequency-amplitude independently controlled neural oscillators in the propeller, is carried out through adjacent coupling, forming a CPG network model for the propeller. Moreover, the output signals of each oscillator in the control model and the influences of parameters on output signals are analyzed by simulation in the propulsion-control modes of symmetric waveform, asymmetric waveform and circumferential waveform respectively. Furthermore, influences of waveform parameters on swimming speed and turning velocity are tested in experiments. The result shows a certain stability and maneuverability of the prototype. Also its straight-line swimming velocity and standing turning velocity increase with both undulating frequency and amplitude, and can be up to 109 mm/s and 93°/s respectively at maximum. Simulation and experimental results prove the feasibility and validity of the CPG control model.