In life science, biological networks usually describe the complex biological processes systematically. Modularity is one of the basic properties of many complex networks. So, biological network modularity is an important research subject of the system biology. Here, we propose a new method to measure the modularity of these networks, and apply it on the yeast transcriptional factors regulatory network and the metabolic networks of several organisms. The results show that the former is not of functional modularity while the latter’s functional modularity is significant. And the application in the metabolic network describes that our index is more suitable for the functional modularity of networks than ACC (Average Cluster Coefficient). In addition, it is effective to do the comparisons between the modularity of differential biological networks or differential module identifying methods by using this method.