Multipactor occurrence essentially depends on the secondary emission property of the surface material, which is, thus, the requisite input for multipactor threshold prediction using the numerical and theoretical approaches. However, secondary emission yield (SEY) deviation in experimental measurements inevitably leads to uncertainty error in multipactor threshold prediction. Therefore, this paper presents a thorough quantitative analysis of multipactor threshold sensitivity to SEY including the effect of the device geometry, the multipactor mode, and the material type. Based on the statistical modeling, multipactor threshold voltages with respect to the SEY variation in critical SEY regions are calculated for both the parallel plates and coaxial lines with different multipactor orders and typical materials. Furthermore, the distribution of electron impact energy is also obtained to elucidate the underlying mechanism for the relevant sensitivity discrepancy. The result reveals that multipactor threshold is generally most sensitive to the energy region below the first crossover energy ( E1), and this is changed to higher energies below the corresponding energy to the SEY maximum ( Em) with a change in the device geometry, multipactor mode, or coating material. It is also found that the magnitude relation of the threshold sensitivity between different regions is radically determined with the distribution of electron impact energy, and the SEY variation close to Em merely affects the threshold result with a high multipactor order. This research provides useful reference for properly determining the threshold margin from the measurement error of SEY, thus promoting the performance optimization with multipactor prevention in the practical application of microwave devices.