Sheep are one of the livestock that can provide benefits such as meat, milk, and skin which are used as a source of income for the community, especially in the Asahan regency area.  However, still, the farmer's knowledge in terms of care and knowledge of sheep diseases can result in the sheep being able to die.  The purpose of this study was to create an expert system to diagnose sheep disease using the theory of Bayes. The waterfall is used as a model in this study with stages, analysis, design, implementation, and testing.  analysis was carried out to find the required data using observations and interviews.  The design of this system consists of Unified Modeling Language (use case ), diagram classes, flowcharts, and interfaces. Testing the system uses a black box which aims to find out the functionality of the system that has been implemented, and data analysis using the theory of Bayes.  Our findings resulted in bayes having an expert system for the diagnosis of diseases in sheep.  The results of Bayes analysis with the theory of Bayes showed that the disease experienced in sheep is deworming with a probability level of 60.71%.  The results of the black box test show that the functionality of this system is already running well