This work is the second in a series of manuscripts devoted to the investigation of the spectrophotometric properties of Saturn’s rings from Cassini-VIMS (Visible and Infrared Mapping Spectrometer) observations. The dataset used for this analysis is represented by ten radial spectrograms of the rings which have been derived in Filacchione et al. (2014) by radial mosaics produced by VIMS. Spectrograms report the measured radiance factor I/F of the main rings of Saturn as a function of both radial distance (from 73500 to 141375 km) and wavelength (0.35–5.1 µm) for different observation geometries (phase angle ranging in the 2°–132° interval). We take advantage of a Monte Carlo ray-tracing routine (Ciarniello et al., 2014) to characterize the photometric behavior of the rings at each wavelength and derive the spectral Bond albedo of ring particles. This quantity is used to infer the composition of the regolith covering ring particles by applying Hapke’s theory. Four different regions, characterized by different optical depths, and respectively located in the C ring, inner B ring, mid B ring and A ring, have been investigated. Results from photometric modeling indicate that, in the VIS-NIR spectral range, B ring particles are intrinsically brighter than A and C ring particles, with the latter having the lowest albedo, while the single particle phase function of the ring’s particles is compatible with an Europa-like or Callisto-like formulation, depending on the investigated region. Spectral modeling of the inferred Bond albedo indicates that ring spectrum can be reproduced by water ice grains with inclusion of organic materials (tholin) as a UV absorber intimately mixed with variable amounts of other compounds in pure form (carbon, silicates) or embedded in water ice grains (nanophase hydrated iron oxides, carbon, silicates, crystalline hematite, metallic iron, troilite). The abundance of tholin decreases with radial distance from C ring (0.2–0.6%) to A ring (0.06%) for the selected regions. Its distribution is compatible with an intrinsic origin and is possibly related to the different plasma environment of the different ring regions. The identification of the other absorber(s) and its absolute volumetric abundance is uncertain, depending on the adopted grain size and mixing modality (intraparticle or intimate). However, assuming a common composition of the other absorber in the ring plane, we find that its abundance anti-correlates with the optical depth of the investigated regions, being maximum in the thinnest C ring and minimum in the thickest mid B ring. In the case of the C ring, an additional population of low-albedo grains is required to match the positive spectral slope of the continuum in the 0.55–2.2 µm interval, represented by an intraparticle mixture of water ice and a spectrum similar to troilite or metallic iron. The distribution of the darkening compounds is interpreted as the result of a contamination by exogenous material, which is more effective in the less dense regions of the rings because of their lower surface mass density of pure water ice.