Volatile flavor compounds from milk were analyzed and identified by using the analysis methods of headspace solid phase microextraction gas chromatography/mass spectrometry (HSPME-GC/MS) and electronic nose (E-Nose) system. About 30 volatile compounds were identified by HSPME-GC/MS for the fresh and off-flavor milk samples. Also, the correlation between rancidity and ageing days of milk was obtained by the aid of principal component analysis algorithm. It shows that the E-Nose system can identify the various types of milk flavor. These results imply that the analysis method based on the E-nose system can apply to the quality control of milk flavor and the rancidity.