Abstract We examine an extensive synthetic aperture radar (SAR) data set from the Arctic Ocean spanning a time period from January to June 2015, with the aim of identifying multi-polarization parameters that can be used to accurately separate newly formed sea ice from the surroundings. Newly formed sea ice areas both provide favourable routing for ship traffic, and are key to Arctic climate science because they enable heat exchange between the ocean and the atmosphere. Our data set encompasses three different frequencies, X-, C- and L-band, at a range of incidence angles, and were acquired under different environmental conditions. Our results suggest that by combining the scattering entropy and the co-polarization ratio we can successfully separate the newly formed sea ice from open water and thicker sea ice within all three frequencies throughout the winter and spring season. We observe a high correlation between scattering entropy values calculated using quad-polarization C- and L-band data and scattering entropy values calculated using the same scenes reduced to the co-polarization channels (HH and VV). We therefore conclude that dual-polarization (HH and VV) X-band scenes can be directly used to complement quad-polarimetric C- and L-band scenes for studies of newly formed sea ice. To confine the quad-polarimetric data sets to their co-polarization channels one can ensure a higher signal-to-noise ratio. Incidence angles below 35° are needed to keep the signal-to-noise ratios sufficiently high for the scattering entropy and co-polarization ratio. Due to its lack of incidence angle dependency, the polarization difference can provide additional support in newly formed sea ice studies. The regular coverage of the Arctic Ocean with C-band SAR means that such scenes should to be included in any automatic monitoring, however, X- and L-band SAR can, based on their difference in penetration depth, provide additional information about newly formed sea ice types and surface structure.