While Saturn's main airless moons are all composed largely of water ice, their respective thermal histories and near environments have led to different regolith compositions and structures. Part of this history is recorded in their subsurface which can be probed by microwaves. Using a combined thermal and radiative transfer model, we here investigate all distant observations acquired in the passive mode of the RADAR on board the Cassini spacecraft (2004–2017) at 2.2-cm wavelength. The joint analysis of the derived disk-integrated emissivities and published radar albedos provides new insights into the purity and maturity of the regolith of Saturn's icy moons. We find that satellite-to-satellite variations and large-scale regional anomalies in microwave signatures primarily reflect different degrees of contamination of the regolith by non-ice compounds. To a lesser extent, they may also point to different concentrations of scatterers in the subsurface; these scatterers must be made of ice and/or void rather than of non-ice contaminants. Enceladus appears to have the cleanest regolith likely due to the geological youth of its surface. Observations also suggest that the current heat flux emanating from this moon is not confined to the South Pole Terrain. In the inner system, the degree of purity of the satellites' regoliths decreases from Enceladus outward likely due to the decrease of the E-ring influx. In the outer system, Phoebe's ring mantles Iapetus' leading hemisphere with a decimetric layer of optically-dark and microwave-absorbent dust. Dione is surprisingly less radar-bright and more emissive than expected from both the observed general trend and the current understanding of its geological history. Another question remains outstanding: why are Saturnian moons, and to a lesser extent Jovian moons, so radar-bright at centimetric wavelengths? Current models assuming purely-random scattering in their subsurface fail to simultaneously reproduce active and passive microwave observations, especially for Saturn's inner moons. This may be due the presence of organized and especially efficient backscattering structures in their subsurface. The challenge is now to identify structures that are geologically plausible.