This project developed a neural network based on nursing system, which a computer application program the network retrieves appropriate noising diagnoses among 98 nursing diagnoses Korea Noises Association recommended when related factors and clinical signs are entered. This nursing diagnosis system consists of five sub-systems: data-input system from which a nurse can obtain knowledge from experts; data-base system in which nurses can manage statistical data; inference-support system which provides learning effect; inference system which increased inference ability of nurses using existing data and nursing diagnoses data; and result-retrieve system in which nurses can look at the outcomes of nursing process. This nursing diagnosis system provides a interface which enables maintenance and revision easier as well as to induce a diagnosis through the communication between the nurse and the patient. In addition, this system makes nurses determine the nursing diagnoses in a more accurate manner by linking the learning effect of the system with intra-neural network algorithm. This nursing diagnosis system will assist nurses to provide foster and more accurate way to implement nursing diagnosis.