Because the visible and near-infrared reflectance spectrum is easily to be acquired using cheap instrument and suitable for online analysis, the features and generation principle of spectral reflectance curve of coal in the visible and near-infrared range were studied and analyzed in this paper. 12 typical types in the three major coal types of anthracite, bituminous and lignite were collected from coal mines in Shanxi, Shandong, Ningxia and Jilin. With decreasing coal rank, these coals included No. 1 anthracite, No.2 anthracite, meager coal, meager lean coal, lean coal, coking coal, fat coal, 1/3 coking coal, gas fat coal, gas coal, No.1 lignite and No.2 lignite. Spectral reflectance curves of these lump coals were acquired in the visible and near-infrared range by a field spectrometer in the laboratory. By analyzing features of these spectral curves, it was found that reflectance curves of anthracites tend to be horizontal in the whole wavelength range and the absorption valleys are not obvious. With decreasing coal rank, spectral reflectance, spectral slope in the near-infrared range, the numbers of obvious absorption valleys and the absorption intensities all increased. Bands of 13 obvious absorption valleys were selected. Carbon structures and mineral compositions of these coal samples were measured through X-ray diffraction (XRD) analyses. With increasing coal rank, the aromatization tendency of amorphous molecular structures of coal plays a major role in reducing the spectral reflectance and flatting the reflectance curve. With decreasing coal rank, overtones and combinations in the near-infrared range generated by the fundamental frequencies of absorption groups mainly including aliphatic side chains in the mid-infrared range generate a lot of absorption superposition. And most of the absorption valleys are not obvious due to the absorption superposition and the relatively more pronounced absorption valleys appear near 1 700 and 2 300 nm. At the same time, transition metals mainly Fe-contained minerals, H2O, clay minerals and other inorganic components are also the factors that increase the number of absorption valleys of reflectance curve of coal. Through X-ray fluorescence (XRF) and proximate analyses of these coal samples, the contents of major mineral elements such as Fe and Al and the contents of moisture, ash, volatile, and fixed carbon based on air-dried basis were acquired. It was shown that the spectral slope of reflectance curve of coal in the near-infrared range is positively and negatively correlated with volatile yield and fixed carbon content respectively. The sum of absorption depths of H2O bands is well linearly correlated with intrinsic moisture content. There is a basically linear relationship between Fe or Al content and the sum of absorption depths of the relevant bands. However, there is a poorly linear relationship between volatile yield and the sum of depths of 1 700 and 2 300 nm absorption valleys which are mainly caused by overtones and combinations of the organic fundamental absorption bands. The acquisition of reflectance spectrum features of typical lump coals in the visible and near-infrared range provides the basis for hyperspectral remote sensing of coal mine areas and establishment of spectral library of coals, and also the reference for the rapid, low cost and qualitative identification of coal categories by shapes of spectral reflectance curves directly. At the same time, it is of great significance to the development of spectral sensor for coal detection in coal mine.