Pulmonary Tuberculosis is a disease that should be recognized and alerted early on by everyone. Pulmonary tuberculosis is a dangerous infectious disease caused by the Mycobacterium group Expert system is made to transfer the expert knowledge into the system, so it hopefully can give solution to problem. “Expert system for diagnosis of infection TBC lung disease” is an alternative for the various expert system that can be used to overcome the problem. The aim of designing expert system is for diagnosis of infection TBC lung disease, should that users can diagnose towards this illnes based on symptoms put into this system. Development of expert system uses knowledge base stored in database using Paradox, rule based with method of dempster shafer and bayes theorem. The result of consultation with the system indicate that this system can determine disease and suggestion based on symptoms inputed by user.