The chest CT(computed tomography) is an important diagnostic method in detecting various pulmonary diseases such as cancer and tumour. In this paper, we present an automatic method that detects solitary pulmonary nodules in two-dimensional chest CT images using image segmentation and feature analysis. First, we segment the lung region from chest CT images, extract candidate nodules from the region, and finally determine real solitary pulmonary nodules among the candidates. The experiment shows that the proposed method achieves considerable performance in detecting solitary pulmonary nodules. It is expected that, by improving the proposed method systematically and integrating it with commercial diagnostic equipments in the future, it can be utilized in real environment for more effective diagnosis.