In order to realize the computer-aided diagnosis, an efficient algorithm is proposed to segment lung parenchyma in CT images. This paper gets the optimal threshold via the means of obtaining minimum between peaks on histogram image. On this basis, using the spatial filtering and connected components labeling to remove interference, getting the lung region at last. The experimental results indicate that the method can effectively segment the pulmonary parenchyma, and solves the problem of preprocessing issue when detecting pulmonary nodules.