Biomimetic robot fish, as a new style of submarine thruster, can play an important role in many applications. Research on information processing and control of the robot fish is mainly carried out in this paper. Biomimetic robot fish’s control, information processing and future applications are studied in different layers, such as bottom control, top planning, coorperation and node design of underwater sensor networks, etc. Researc productions are as follows. Firstly, the application background and research purposes of biomimetic robot fish are introduced. The research trend and progress in fish’s propulsive mechanism, robot fish’s research, development and control in the world are reviewed. The background and structure of the thesis are also given. Secondly, autonomous reactive control for obstacle avoidance of the robot fish is studied. The design of a biomimetic robot fish system with ultrasonic and infrared sensors is given. An obstacle avoidance strategy based on reinforcement learning is proposed in this paper. The state and behavior sets are designed. Through adopting immediate and delayed rewards, valid state-behavior pairs are gotten through learning. Thirdly, an adaptive image segmentation algorithm is put forward in this paper. The proposed algorithm can improve the segmentation performance, and solve the problem that image segmentation algorithm based on fixed thresholding cannot suit the light variation. The algorithm is applied to the auxiliary visual system for the biomimetic robot fish control. Fourthly, based on visual information, a kind of Bezier curve-based path generation and motion control algorithm for the biomimetic robot fish is proposed in the paper. A smooth Bezier curve can be planned from robot fish current state to the expected state. Then multiple discrete points are gained from the curve. Robot fish’s locomotion along the planned path is realized through direction and velocity control. The validity of the algorithm is tested through a shooting gate experiment. Fifthly, coorperation algorithm of multiple robot fishes for object searching is proposed. Focusing on a practical problem, finding the heat source position in finite water area, the algorithms of forming equilateral triangle and searching based on formation technique are presented to locate the heat source position. Sixthly, the architecture for underwater mobile sensor networks is presented in this paper. The robot fish acts as the node in this architecture. An underwater monitoring system based on a robot fish is built up. This system is composed of robot fish subsystem, communication subsystem and controller. Finally, the obtained results are summarized and future work is addressed.