A rotor system is composed of a rotating shaft with supporting bearings. The rotor system is widely used in every rotating machinery such as the turbine generator and the high precision machine tools. A negligible error or malfunction in the rotor, however, can cause a catastrophic failure in the system then result in the environmental and economic disasters. A diagnosis of the rotor system is important in preventing these kinds of failures and disasters. Up to now, many researchers have devoted in the development of diagnosing tools for the system. The basic principles behind the tools are to retrieve the data through the sensors for a specific state of the system and then to identify the specific state through the heuristic methods such as neural network, fuzzy logic, and decision matrix. The proper usage of the heuristic methods will enhance the performance of the diagnostic procedure when together used with the statistical signal processing. In this paper, the methodologies in using the above 3 heuristic methods for the diagnostics of the rotor system are established and also tested and validated for the data retrieved from the rolling element bearing and journal bearing supported system.