Spectral imaging technology, widely used in remote sensing applications, such as satellite or radar imaging, has recently gained importance in the field of artwork conservation. In particular, multispectral imaging in the near-infrared region (NIR) has proved useful in analyzing ancient paintings because of the transparency of most pigments and their varied reflectance changes over this spectral region. A variety of systems, with different detectors and filtering or dispersing technologies, have been implemented. Despite the recognized potential of multispectral NIR imaging, which provides information on both spectral and spatial domains (thus extending the capabilities of conventional imaging and spectroscopy), most of the systems currently used in art diagnostics have limitations. The technology is still in its early stages of development in this field. In this Account, we present the scanning multispectral IR reflectography (SMIRR) technique for artwork analysis, together with an integrated device for the acquisition of imaging data. The instrument prototype is a no-contact optical scanner with a single-point measurement of the reflectance, capable of simultaneously collecting a set of 14 spatially registered images at different wavelengths in the NIR range of 800-2300 nm. The data can be analyzed as a spectral cube, both as a stack of wavelength resolved images (multi-NIR reflectography) and as a series of point reflectance spectra, one for each sampled pixel on the surface (NIR spectrometry). We explore the potential of SMIRR in the analysis of ancient paintings and show its advantages over the wide-band conventional method. The multispectral option allows the choice of the most effective NIR bands and improves the ability to detect hidden features. The interband comparison aids in localizing areas of different pictorial materials with particular NIR reflectance. In addition to the analysis of single monochromatic images, enhancement procedures involving the joint processing of multispectral planes, such as subtraction and ratio methods, false color representation, and statistical tools such as principal component analysis, are applied to the registered image dataset for extracting additional information. Maintaining a visual approach in the data analysis allows this tool to be used by museum staff, the actual end-users. We also present some applications of the technique to the study of Italian masterpieces, discussing interesting preliminary results. The spectral sensitivity of the detection system, the quality of focusing and uniformity of the acquired images, and the possibility for selective imaging in NIR bands in a registered dataset make SMIRR an exceptional tool for nondestructive inspection of painting surfaces. The high quality and detail of SMIRR data underscore the potential for further development in this field.