Abstract: Deeper knowledge of social networks' structure and temporal evolution enhances data mining for both research and education purposes. An empirical analysis of a Chinese social network, Renren, shows that it follows an exponentially truncated power law in degree distribution, and has a short average node distance.