In recent times, the use of natural ventilation in office buildings, schools or public facilities has become popular. The purpose of natural ventilation is not only to reduce air conditioning usage and enhance comfort but also to facilitate the improvement in intellectual productivity. However, natural ventilation rate must be appropriate in order to achieve these results. Since natural ventilation rate is affected by changes in outside conditions, to execute natural ventilation control, it is necessary to adjust the opening area. Therefore, develop simulations that can help evaluate natural ventilation control. The main purpose of this study is to evaluate the performance evaluation of natural ventilation control in office buildings. In this study, we first surveyed the actual state of buildings with natural ventilation. In addition, following the survey, we examined the basic characteristics of the buoyancy driven ventilation using airflow network analysis. In the survey, the elements that affect the natural ventilation rate were identified. We surveyed the effective opening area of natural ventilation opening (αAi), natural ventilation path, and the movement of buildings that use natural ventilation. We found three things: 1) αAi is correlated with typical floor area, 2) about 85% of the naturally ventilated buildings have adopted the buoyancy driven ventilation, 3) there is a recent technique to separate the chimney for effective natural ventilation. The details of this survey are given in Chapter 3. We carried out three numerical analyses of natural ventilation rate and neutral pressure level due to change αAi. These analyses assumed ten-story buildings. 1) We calculated air changes per hour due to opening and closing of the entrance. There was not any significant effect on the natural ventilation rate of the second floor and more due to a sufficient increase in the neutral pressure level. 2) We suggested two methods to reduce the difference between the natural ventilation rates of each floor. First, we calculated air changes per hour due to a smaller αAi on the lower half of the floor. Second, we calculated air changes per hour due to separate chimneys into two, upper and lower, of the floor. According to these analyses, αAi of the lower half of the floor can be set to 50% of the αAi of the upper half of the floor or the chimney can be lowered up to sixth floor to reduce the difference between natural ventilation rates each floor. 3) We determined the neutral pressure level. In brief, there are three findings. First, αAi on all the floors should be same. Second, αAi on the lower half of the floor should be small. Finally, the chimney's protruding length set to be longer. Following these suggestions, one can determine neutral pressure level. Furthermore, we presented a graph for determining αAi to achieve appropriate natural ventilation rate, which is called “Isodose curve of air changes per hour.” Details of these analyses are given in Chapter 4. In future research, the proposed airflow network analysis method can be used to evaluated the performance of natural ventilation control in office buildings.