Abstract: The safe operation of the large-medium hydraulic turbine set has great influence on the hydropower station and the grid. With analysis of the principle of the condition monitoring and fault diagnosis of hydraulic turbine, a monitoring and diagnosis system for hydraulic turbine set was developed based on novel networks. The system had three layers that were interconnected by the RS-485 bus and Ethernet. The field nodes were designed based on 8-bit AVR ATmega8515. The monitoring station used a high performance industry PC to monitor the status of the hydraulic turbine, and had the functions of fault diagnosis by wavelet analysis and neural network method. The internet based remote monitoring and diagnosis system for the turbine set was realized by the web server. The experiment results showed that the vibration signals were measured accurately, and the two common types of turbine faults could be recognized effectively.