Recently, increasing attention has focused on exploring brain activity during rest. A current challenge is to ascertain activity patterns without using an external reference. In this paper, we propose a model-free approach based on regional homogeneity (ReHo) to identify specific network of brain regions during rest. In the context of the ReHo, it seems to be reasonable to assume that the voxels within a functional cluster have higher temporal coherence than that within a general cluster. Using this technique, a large distributed network during rest was identified and compared with previous meta-analysis studies. The proposed method shows an extensive ability in analyzing resting state data.