Expertise systems represent one of the most important sciences of artificial intelligence that offer systems developed to take the place of experts, especially in a specific field, where these systems can be consulted to perform a specific task in the specialized field of this system. Expert systems play an effective role in the scope of medical diagnosis in terms of identifying sites of weakness and causes of deficiencies in cases that are difficult to diagnose based on the symptoms observed from the diseases. The expert system is used as an intelligent medical diagnostic based on patient data. This study aims to develop an expert system to diagnosis thyroid diseases using knowledge representation of production rules. The knowledge base and the inference mechanism has discovered and included in the system, it depends on the forward inference method in building the diagnostic system. Thyroid disease is considered one of the most dangerous and widespread diseases due to its severity as a fatal disease that make it difficult to diagnose types of thyroid diseases. This study presents the medical diagnostic system for thyroid diseases of the five most common types, Hypothyroidism, Hyperthyroidism, Hashimoto, Graves and Nodules thyroid. The system was tested and evaluated and its performance by presented to experts and specialists in medicine field with 85% diagnostic accuracy on 92 cases. It can be utilized to diagnose thyroid diseases with faster, accurately and high efficiency than the traditional diagnosis