WHO in 2013 released data in the Global Tuberculosis Report that 8.6 million more people were positive for TB, and 15.12% of them died. One of the causes of death due to tuberculosis is too late to know when it is infected and too late to get treatment. For an initial diagnosis of whether someone has TB, they can immediately go to a hospital or use a computer application that has an expert's ability. This study aims to design an expert system for the early diagnosis of intestinal tuberculosis using the Forward Chaining method. Expert System is one part of Artificial Intelligence that can create a computer program to provide decisions and analysis like an expert or expert in a particular field. The research method used in this study is the Research and Development method with the ADDIE approach. The expert system for early diagnosis of intestinal tuberculosis is designed by making a decision tree diagram to find out the rules needed to diagnose whether a person has intestinal tuberculosis from the symptoms experienced. The formulation of the problem or the rules needed for the initial diagnosis of intestinal tuberculosis in this study was built using the Forward Chaining method. The output of this research is an Expert System in the form of a Stand-Alone application that is built with the Visual Basic Programming Language, which can provide diagnostic results whether a person suffers from intestinal tuberculosis from the symptoms entered into the Expert System Interface.