In this paper, a level set approach has been developed for segmentation of retinal vasculature. There have been a few studies for segmenting of blood vessels in the retinal images based on the level set method in the literature. Because of the structure of blood vessels, it is difficult to develop such an approach. However, recently many level set methods have been developed to overcome such problems. In this study, an approach based on an augmented Lagrangian method that produces faster and more accurate results for the segmentation of blood vessels has been proposed. The accuracy of the presented method was also supported by a visual result.