The visible-near infrared spectra of minerals and rocks are closely related to their intrinsic physicochemical properties. The determinants of the reflectance spectra are the composition and the content. The study is focused on the relationship between reflectance spectrum and the grade of Anshan iron mine. The visible-near infrared spectrum of Anshan iron mine with different grade was obtained by SVC HR-1024. The influence of grade on reflectance spectra was analyzed. The sensitive band between the grade and the spectra of the samples were determined. The model for predicting grade of experimental samples was established. The results show that, the effects of grade on the spectra of hematite and magnetite are different. The grade of hematite not only affects the reflectance spectra of samples, but also affects the shape of the reflectance spectra of samples. The grade of magnetite does not affect the shape of reflectance spectra of samples. It only affects the reflectance spectra of samples. The influence of hematite grade on reflectance spectra is not uniform at different bands. In 350~1 000 nm band, the reflectivity of samples is sensitive to grade. There is a significant linear negative correlation between reflectance and grade of hematite in this band. In 1 000~1 250 nm band, the slope of reflectance spectrum is sensitive to grade. There is a significant linear positive correlation between them. However, in 1 250~2 500 nm band, the reflectance spectra are slightly affected by grade. The quantitative relationship between the reflectance spectra and the grade of magnetite was analyzed. There is a significant negative correlation between grade and reflectance with the exponential function. The relationship is almost the same in the 350~2 500 nm band. Then, the grade inversion model of hematite and magnetite had been established. We tested and verified the practicality of the model. The error of these models is less than 1%. The prediction results of these models are ideal. In this paper, the influence of grade on reflectance spectra was clarified. The inversion model of the iron grade was established. It provides a new method for determination of iron grade by spectral analysis technology.