Rapid qualitative and quantitative detection for melamine in pure milk was studied by using near infrared spectra(NIRS). Experiment was conducted by preparing two groups pure milk samples in which melamine content was different for qualitative and quantitative analysis. By combining NIRS technology with the cluster analysis method, two kinds of milk samples with and without melamine were classfied effectively. Spectrum pretreatment and wavelength choice methods were employed before model optimization. In this way, NIRS quantitative model for detecting melamine was established. The results show that NIRS model of detecting melamine content in pure milk has good stability and predictive ability. It suggestes that NIRS can be used as a quick, green and convenient method for detecting melamine in dairy.