The use of microwave radiation data collected by lunar orbiters to infer the lunar brightness-temperature (TB) distribution is of great importance to lunar scientific exploration. The TB data acquired by China's Chang'E-2 passive microwave radiometer (MRM) is one of the basic materials for lunar temperature mapping. Core issues of lunar microwave TB mapping include data fitting at different latitudes, outlier removal, and integrated spatial interpolation considering influencing factors. We applied a third-order Fourier function to fit the lunar TB data, which effectively removes outliers considering the fitting characteristic and three-sigma rule (FCTSR). Because the lunar surface temperature is influenced by a variety of factors such as topography and albedo, we then built a multi-factor co-kriging interpolation method to perform lunar TB mapping accurately. The cross-validation shows that, in terms of the mean-absolute-error and root-mean-square-error, the multi-factor co-kriging interpolation improves the mapping accuracy by 30%–60% compared to inverse distance interpolation and ordinary kriging interpolation. The analysis shows that topography and albedo are the most important factors influencing the lunar TB, and the TB maps at 37 GHz reveal cold spots that can be considered as younger craters. It is concluded that the proposed data fitting, denoising and spatial interpolation methods significantly improve the lunar TB mapping. The results and scientific data can also provide a basic energy map for lunar roving path planning and subsequent lunar exploration. •TB outliers were effectively removed using the fitting characteristics and three-sigma rule.•A multi-factor co-kriging method was developed to produce accurate lunar TB maps.•The co-kriging improves accuracy by 30%–60% compared to IDW and ordinary kriging.•The TB maps at 37 GHz reveal cold spots that can be identified as younger craters.