Annular mould is a primary part of the feed pellet mill machine, and drilling is a key processing for the CNC manufacturing of annular mould. The operating mode noise of drill bits is tested in this paper, in order to distinguish the wear condition of drill bits by the noise analysis on power spectrum. By using the computer, embedded system and short distance wireless communication technologies, a novel drilling noise monitoring system and corresponding and a humancomputer interaction software are achieved for the CNC drilling machine, to distinguish and dispose the wornout drill bits automatically. The test result shows that the power spectrum characteristic of drill bit noise is distinct, as well as the monitoring system for wear condition of drill bits is reliable, and it has been applied and test in actual productions.