Many of the industries depend on the application of the expert system in diagnosis field; to gain the benefits and to avoid stopping or tripping their industries. Availability of the good diagnosis system gives a clear picture of the fault type with recommendations and advice like the human thinking. The human thinking that is represented by experienced people cost money to bring them or even to ask them. At the present, people who work with MSF desalination plant usually perform MSF troubleshooting diagnosis using their field experience and vendor troubleshooting manual knowledge. They try to relate the probability of the occurrence of a symptom to an underlying cause. However, this requires specific knowledge. With units becoming more complex, the diagnostic knowledge become more difficult to process by ordinary operators. Therefore, there is a need for an expert system to give its knowledge and advice to the operators. This project was designed to develop diagnosis or troubleshooting actions required in the operation of the multi stage flash desalination. The work was based on using fuzzy logic approach to generate the troubleshooting information. This can help the new trainee operators to know what are the required actions to be done when a certain problem happens. In addition, this program will assist the operators in the MSF desalination plant to take a fast and reliable action whenever an operation problem occurs so the actual cause can be identified and fixed as soon as possible. Most applications of fuzzy logic were in control field. This present work gave detailed information about the previous work about the application of the fuzzy logic in the diagnosis system or as an expert system. The study covered detailed information about multi stage flash desalination and its operation. A questionnaire (i.e., survey) was distributed among thirty experienced operators working in the multi stage flash desalination plant. They were asked about their actions if they face certain troubles or faults during the operation of the MSF. All the feed back from the expert people were organised in a certain format to be entered in the fuzzy logic expert system. Then, all the troubleshooting statements were entered to the expert system in term of actions as a base knowledge. The expert system was build for the brine heater that is the main part of the desalination plant in order to check its validity, The inputs were the readings from the sensors and the output was in terms of actions and sound. If any of the input were abnormal then the fuzzy logic-based expert system output would vary in the range [-100 100] according to urgency weight. There would be a sound to indicate the abnormality in the operation. If the input reading were normal then the output of the expert system would not give any actions. The actions or the troubleshooting statements were built. The fuzzy logic-based expert system was tested by using off-line data. The off-line data was introduced to the expert system by using the Matlab interface. The results of fuzzy logic-based expert system were tested and evaluated by the experienced operators. The performance of the expert system was judged to be satisfactory by the human experts. They recommended that the expert system could be used on line without any problem.