AbstractCloud phase is an important factor affecting cloud contributions to the polar energy budget. Brightness temperature differences (BTDs) between two mid‐infrared (mid‐IR) spectral channels are commonly used in satellite remote sensing to determine the cloud phase, but the mid‐IR has limitations for cloud phase determination in polar regions. This study explores the synergy between the far‐ and mid‐IR for ice cloud phase determinations over polar oceans. A far‐IR BTD test (BTD449‐521) is developed based on the spectral variations of both ice cloud scattering and absorption properties in the far‐IR “dirty window” region (400–600 cm−1 or 16.7–25 μm). Synthetic IR radiances at 1 cm−1 spectral resolution are generated using ERA5 reanalysis data for both polar regions. A subset of these spectra is used to compare the ice phase determination skill of the far‐IR test with an “AIRS‐Like” mid‐IR BTD test (BTD1231‐960) for combinations of effective ice particle diameter (Deff_ice), cloud optical depth (COD), and cloud top pressure (CTP). The far‐IR test is performing better for ice clouds with the smallest Deff_ice that we have studied (20 μm) due to the sensitivity of far‐IR scattering to ice particle size. The mid‐IR test was either comparable to or more successful than the far‐IR test for ice clouds with large particle sizes. For all Deff_ice, increasing COD leads to enhanced far‐IR BTD signals due to stronger multiple scattering in the far‐IR than in the mid‐IR. Overall, the variations in far‐IR and mid‐IR BTD test performance are strongly sensitive to Deff_ice, followed by COD and CTP.