In a telecommunication system, call detail records (i.e., CDRs) are generated automatically for tracking and billing purposes when mobile users having calls. To further investigate the information buried in huge amounts of CDRs, relationship among mobile users can be organized. Specifically, communities of acquainted mobile users can be effectively discovered from collected CDRs through our approach proposed in this paper. Note that understanding the communities and corresponding calling behaviors are of great importance to telecommunication companies. To conduct proper community mining on CDRs, techniques of data transformation and social network analysis are fully exploited. Our study shows that the proposed approach is practically feasible.