Borehole-based subsurface electromagnetic (EM) measurements, namely, galvanic resistivity (laterolog), induction, propagation, and dielectric dispersion logs, are commonly used for water-saturation estimation in hydrocarbon-bearing formations. EM logs exhibit frequency dependence due to the interfacial polarization (IP) effects arising from clay-grain surfaces, conductive minerals, and charge blockage in pore throats. IP effects in shale formations adversely affect the log-derived water-saturation estimates, especially when there is low porosity, high salinity, the presence of pyrite disseminations, and high clay concentration. Conventional EM log-interpretation methods estimate water saturation in shale formations by separately interpreting the galvanic, induction, propagation, and dielectric dispersion logs using various empirical models or mixing laws. This approach leads to significant variations and uncertainties in petrophysical estimations. We have developed an inversion-based joint petrophysical interpretation of multifrequency effective electrical conductivity and dielectric permittivity logs derived from various combinations of the four aforementioned downhole EM logs acquired in clay- and pyrite-rich shale formations. The proposed joint-interpretation method uses a single mechanistic model that accounts for the IP effect arising from clay and conductive mineral grains, thereby generating physically consistent water-saturation estimates in shales. The proposed inversion-based interpretation also generates estimates of formation brine conductivity, surface conductance of clay, and average radius of clay and conductive mineral grains. The proposed method is applied to one field case and three synthetic geologic formations, with varying clay type, conductive mineral properties, and water saturation. Further, the sensitivity of inversion-derived estimates to the presence of various types of noise in the EM logs is investigated. The joint petrophysical inversion algorithm is applied to field broadband dispersion EM data acquired in an organic-rich shale formation. Water saturation, brine conductivity, surface conductance of clay, and radius of clay were consistently estimated in the shale formation using various combinations of available EM logs.