Abstract The Bosch process was studied at a substrate temperature of −100 °C and compared to etchings performed at room temperature, as in the general case. The tests were realized using an inductively coupled plasma reactor by varying C 4 F 8 passivating gas flow injections both at +20 °C and −100 °C. It was observed that the Bosch process is effectively temperature dependent and that the necessary C 4 F 8 passivating gas flow can be reduced to obtain similar anisotropic profiles at −100 °C compared to the ambient temperature process. For example, in one of the studied cases, a fluorocarbon injection of 8 sccm was sufficient to obtain an anisotropic etch rate of up to 4.4 μm min −1 at −100 °C whereas the profile obtained at +20 °C using the same parameters presents lateral etching defects with a reduced etch rate of 2.4 μm min −1 . At this point, the C 4 F 8 flow must be increased to 12 sccm (50% more) to retrieve an anisotropic profile with an etch rate of 4.0 μm min −1 . In the case of cryogenic Bosch (cryo-Bosch) processing, C 4 F 8 feed dosing has a greater influence on the passivation regime which affects the subsequent etching result but it can be easily refined through the optimization of process parameters. An in-situ ellipsometry study of the deposition rate of C 4 F 8 on both polycrystalline silicon (p-Si) and SiO 2 substrates was realized by varying the gas flow at −100 °C and +20 °C. This study shows that the deposited fluorocarbon material is approximately a hundred times thicker at cryogenic temperatures using the same process parameters. Scanning electron microscopy (SEM) observation of these samples are in adequacy with the ellipsometry results. Cryo–Bosch etching also results in a slightly higher etch rate compared to room temperature processing when analyzing similar anisotropic profiles. Si:SiO 2 etching selectivity is significantly increased at −100 °C although the aspect-ratio dependent etching phenomenon is more important.