In the present study, the tool wear has been monitored using the cutting sound acoustic spectrum and the linear predictive cepstrum coefficient (LPCC) of the milling sound signal would be extracted to be used as the acoustic spectrum characteristic parameters. The relationship between each order component of LPCC and the flank wear of the tools was analysed. The experimental results show that there are clear characteristic components in the milling sound signal related to the tool wear. It has been found that the characteristic components associated with tool wear are mainly concentrated in the sixth-, seventhand eighth-order components of LPCC.