Implementation of the nurse scheduling task is a difficult problem domain for head because it is a multi-period, multi-criteria decision-making problem with the condition that it needs to meet the requirements of the patients, the hospital administration, and the staff nurse. There have been many research studies done for nurse scheduling using the following methods: cyclic method, mathematical modeling method, heurestic method and expert system approach. Without a theoretical review of the head nurses' problem-solving process, the limitations of inflexible computational modeling in comparison to the complex, real problem-solving process cannot be overcomed. This research applies the cognitive scientific approach which uses protocol analysis and computer simulation. It will be used as basic data for a decision support system which will be constructed as a result of protocol analysis(collect data, analyze the head nurses' problem-solving process)and cognitive process simulation. Following research will develop the following research lists: first; construction of databases including each nurse's major attributes as a data field, second; development of schedule optimizing modules, and third; development of user interface for communication between head nurses and staff nurses using the system.