For the purpose of developing a method for rapid detection of total nitrogen content in organic vinegar residue substrates, visible-near-infrared spectroscopy was used to collect the spectral data of the representative samples. And the spectra were pretreated by the method of first derivative after normalization. In order to choose the characteristic wavelengths for the total nitrogen content in the substrates, stepwise regression (SWR) method was applied twice. The results indicated that the characteristic wavelength was at 1699 nm, 746 nm, 1864 nm and 2154 nm, respectively. Based on these wavelengths, a linear regression model was performed. For comparing the predictive precision of the model, the correlationship between prediction total nitrogen content and that obtained by chemical method was analyzed. It was confirmed that the multivariate equation was the best with correlation coefficient 0.9334 and root mean squared error of prediction (RMSEP) 1.04. Therefore, it was demonstrated that visible-near-infrared spectroscopy with characteristic wavelengths selecting could be an available method for rapid detection of total nitrogen content in organic substrate.