The space-based radiance standard is of great significance for the study of the satellite observing climate change. The establishment of the space-based radiance standard can not only improve the relative accuracy of satellite observations, but also meet the traceable needs of other satellites through intercalibration. The spectral resolution of the hyperspectral standard remote sensor in space for intercalibration has a significant effect on the intercalibration relative accuracy. The simulations of spectral radiances by modes of MODTRAN and AER LBL are used as proxy data of the hyperspectral remote sensor in reflective solar bands and thermal emissive bands, respectively. The influence of spectral sampling on the observation of spectral radiation and the uncertainty of the radiation standard intercalibration caused by the spectral sampling are analyzed. Considering five kinds of underlying surfaces and six kinds of atmospheric conditions, the difference of spectral radiation under different spectral sampling frequencies is compared, and the spectral uncertainty of the space-based radiance standard intercalibration is evaluated with the utilization of the sensitivity experiment method with MERSI-II as the target remote sensor. The results show that the larger the spectral sampling frequency, the greater the difference in spectral radiation. The maximum radiation difference is up to 100% in atmospheric absorption spectra, low signal spectra, and near ultraviolet solar dark-line spectra. In the atmospheric window, the spectral sampling frequency better than 4 nm can produce radiance to meet the on-obit intercalibration standard with an uncertainty less than 0.3% in reflective solar bands, and the spectral sampling frequency better than 2 cm -1 can also produce radiance temperature to meet the on-obit intercalibration standard with an uncertainty less than 0.1 K in thermal emissive bands. In the near ultraviolet solar dark-line spectra and the atmospheric absorption region, the intercalibration of reflective solar bands is very sensitive to the spectral sampling. The uncertainty of intercalibration is up to 40% at a sampling frequency of 4 nm in the channel with a central wavelength of 1.38 μm. The spectral sampling of 0.8 cm -1 can produce radiance temperature to meet the on-obit intercalibration standard with an accuracy of 0.1 K in the weak atmospheric absorption channel with a central wavelength of 7.2 μm in thermal emissive bands.