Nowadays the need for a safe, economic and reliable power supply is one of the most important factors in Malaysia. For that reasons, Tenaga Nasional Berhad (TNB) try to minimize the power failure in the grid system especially for the impoortant or critical line. From the previous experience, most of the power disturbances are due ti weather, for example lightning. Some of the disturbances cannot be identify from the relay indication. These will cause the fault diagnosis in the power transmission system becomes more difficult and takes longer time to restore back the supply. To avoid this situation, TNB start to introduce a disturbance recorder that can monitor the behaviour of the TNB grid power system and associated equipment. This is to ensure that the fault diagnosis can eb done easily by using the output from the distrubance recorders. A fault diagnosis in transmission system involves identyingthe location and causes of teh faults. This project will discuss a few well -konown fault diagnosis philosophies and fault diagnosis metods that being used by TNB. It will also describe a proposal method to be use in order to overcome the difficulties in conventional fault diagnosis methods. The aproposl method will be a knowledge base computer programming system that been develops in this project. A fault diagnosis method that been used by Asea Brown Boveri (ABB) in their fault diagnosis package will be discuss briefly just for comparison. On top of that a few simulation events will be used to validate the knowledge base fault diagnosis method in this project. Lastly, several real tripping events will be include and analyze for 275kV Serdang's Substation.