Chatter is a complex phenomenon characterized by unstable, chaotic motions of the tool and by strong anomalous fluctuations of cutting forces. The situation becomes more serious in the milling of titanium alloys because of their low Young modulus and extended elastic behaviour. This paper presents an online chatter detection system based on the analysis of cutting forces, which is one of the integrated modules of a multi-sensor chatter detection system consisting acoustic and acceleration sensors. The cutting force is transformed into frequency domain by applying Fast Fourier Transform (FFT). Chatter frequency is identified in the frequency domain by comparing its power spectrum with predefined threshold. Experiments were carried out to validate the mythology.