Abstract. The DC resistivity method is a common tool in periglacial research because it can delineate zones of large resistivities, which are often associated with frozen water. The interpretation can be ambiguous, however, because large resistivities may also have other causes, like solid dry rock. One possibility to reduce the ambiguity is to measure the frequency-dependent resistivity. At low frequencies (< 100 Hz) the corresponding method is called induced polarization, which has also been used in periglacial environments. For the detection and possibly quantification of water ice, a higher frequency range, between 100 Hz and 100 kHz, may be particularly interesting because in that range, the electrical properties of water ice exhibit a characteristic behaviour. In addition, the large frequencies allow a capacitive coupling of the electrodes, which may have logistical advantages. The capacitively coupled resistivity (CCR) method tries to combine these logistical advantages with the potential scientific benefit of reduced ambiguity. In this paper, we discuss CCR data obtained at two field sites with cryospheric influence: the Schilthorn massif in the Swiss Alps and the frozen Lake Prestvannet in the northern part of Norway. One objective is to add examples to the literature where the method is assessed in different conditions. Our results agree reasonably well with known subsurface structure: at the Prestvannet site, the transition from a frozen lake to the land is clearly visible in the inversion results, whereas at the Schilthorn site, the boundary between a snow cover and the bedrock below can be nicely delineated. In both cases, the electrical parameters are consistent with those expected from literature. The second objective is to discuss useful methodological advancements: first, we investigate the effect of capacitive sensor height above the surface and corroborate the assumption that it is negligible for highly resistive conditions. For the inversion of the data, we modified an existing 2-D inversion code originally developed for low-frequency induced polarization data by including a parametrization of electrical permittivity. The new inversion code allows the extraction of electrical parameters that may be directly compared with literature values, which was previously not possible.