The electron cyclotron emission measurable from the outboard side of ITER plasmas is estimated. The effects of harmonic overlap and polarization scrambling are reviewed with the aim of assessing the impact of any polarization change that might occur in the collected radiation before the O and X polarizations are separated. It is confirmed that any polarization scrambling occurring during the reflection at the wall would not alter the measured intensities of lower harmonics of either the O or X mode but would affect only the higher harmonics, which are optically thin. For the second-harmonic X mode, the observed intensity in the 300- to 400-GHz range is considerably lower than that of the O polarization. Hence, this frequency range may be particularly vulnerable to any O-to-X polarization change occurring prior to their separation into different transmission channels. It is shown that if the electron temperature Te near the core is to be measured to within 10% accuracy, the above polarization fidelity should also be preserved to within 10% or better. It is suggested that this requirement may have impact on the location of the polarization splitter unit. Further analysis is required to evaluate the error in the calculation of Te profiles from the measured Trad values arising due to uncertainties introduced by any polarization conversion.