The study on soil profiles and the various soil horizons they consist of are of great significance to the studies on soil genesis and development, soil classification, and some other disciplines of soil science. Traditional soil information acquisition methods are both time-and labor-consuming, however, the proximal soil sensing technology can be used to provide soil information of various scales rapidly and periodically, and has been widely applied to researches such as soil resource survey, land quality evaluation, soil classification, soil mapping, etc.. Traditional methods for measuring soil physical and chemical properties are complicated, time-consuming and costly, and can hardly meet the demands for rapid monitoring of changes in soil property. In recent years, the spectrometric technology has extensively been used to quantitatively analyze samples in a fast, simple and non-destructive way in various fields with results. The data acquired with using the imaging technology combined with the spectroscopic technology are high in both spatial resolution and spectral resolution, and contain very rich soil remote sensing information and hence can provide a solid foundation for quantitative monitoring and mapping of soils properties in the horizontal dimension. However, in the light of the researches done by scholars both at home and abroad, it appears that the study of soil science still lacks an imaging technique high in spatial and spectral resolution, specifically for measuring soil total nitrogen (TN) contents in entire soil profiles. At present, quantitative analyses of soil properties mostly use soil samples collected in the topsoil, 0 ~ 150 mm or 0 ~ 200 mm in depth; studies on soil point samples in the profile 0 ~ 1 000 mm in depth are rarely reported, and little has been found in literature on dot samples in the 0 ~ 1 000 mm soil profile, let alone reports on mapping of vertical distribution of soil TN contents in entire soil profiles. In view of this situation, this study is oriented to explore the feasibility of using the technology of vis-NIR imaging spectroscopy in instant prediction of vertical distribution of TN contents in soil profiles. A total of 3 soil profiles (0 ~ 1 000 mm) were collected from Chongyang, Hubei Province, China. Vis-NIR hyperspectral images (753 spectral bands in 400 ~ 1 000 nm) of the profiles were taken with an imaging spectroscoper quipped with a 25 μm slit, a 35mm focus lens 13.1° in angle of field and an area array CCD of 1 004 pixels × 1 002 pixels. First, the digital photos with fixed scenical grid scale taken by a digital camera were put to undergo geometric correction with reference to the hyperspectral images of the profiles to solve the problem of image deformation caused by technological limitations of the spectrometer and the shooting platform, and then modified to 1mm in image precision. Through spatial and spectral dimensional clipping, pixels of the wooden frame and platform background were removed leaving only soil image data (160 pixels×980 pixels) and valid spectral bands (470 ~ 1 000 nm). After geometric correction and clipping, the images were processed with a variety of supervised classification methods. Results show that the minimum distance method is the best at distinguishing invalid data (e.g., shadows and cracks) from soil data. A "sampling panel" method was proposed for strip-sampling with the panel in line with the specified, averaging the samples similarly to ROI and finally solving the problem of scale inconsistency between point samples and profile spectra. Furthermore, a PLSR calibration model was built up based on the spectral data of 10 spot soil samples, and used to predict TN contents in three complete soil profiles based on their spectral images. Results show that the technology of vis-NIR imaging spectroscopy could be used to inverse and map soil TN vertically in profiles with good prediction results. Verification with measured data demonstrates that R² and RPD was 0.56 and 1.41, respectively, for the 0 ~ 1 000 mm soil layer, which indicates that the prediction method reached the range of rough estimation. And for the 0 ~ 600 mm soil layer, the effects were better with R² and RPD being 0.87 and 1.76, respectively, which indicates that the technology of vis-NIR imaging spectroscopy might have some limitation in the vertical direction, though it can well restore the soil TN distribution patterns in the topsoil layer, especially in the 0 ~ 600mm soil layer. The above findings demonstrate that this study has preliminarily established a set of procedures for soil TN inversion and mapping using the technology of vis-NIR spectroscopy, and the method is applicable to rough estimation of soil TN contents in whole soil profiles.