Abstract: The study aims to search for the hyper-spectral bands most sensitive to soil moisture content and their inversion model. Taking black soil in Jilin Province as the study object, the spectra of soil samples with different soil moisture content were measured by ASD FieldSpec FR at the indoor environment. Soil hyper-spectral characteristics were analyzed. The hyper-spectral reflectance data were transformed to several spectral indices to analyze the relationship of soil moisture and extract sensitive bands. Quantitative inversion model of soil moisture content was carried out by using the stepwise multiple linear regression and exponential analysis. Results showed that the sensitive bands of black soil spectral reflectance (R) and its transformation of the first derivate differential (R′) and logarithm of the first derivate differential (lgR)′ mainly focused in 400-410, 1?400-1?850 and 2?050-2?200 nm when the soil moisture was below field water holding capacity. The highest correlation coefficient between laboratory spectral data and soil moisture reached to 0.89 at 2156 nm. At bands of 1?328, 1?439, 1?742 and 2?156 nm, the prediction model of black soil moisture content was the best when using spectral reflectance logarithm of the first derivate differential for estimation. The coefficient of determination was 0.931. The hyper-spectral inversion model of black soil moisture content provides a new approach for rapid soil moisture monitoring.