The spectral consistency of infrared focal plane detector is one of the important parameters to evaluate the level of material preparation. The detector materials are different because of different preparation processes. Due to the limitations of the speckle size of the spectrometer and the dewar structure of the detector, as well as the huge amount of spectral data collected by the device, it is difficult to test and analyze the detector components. A new method is proposed in this paper. Two detector samples prepared by traditional and optimized process are selected, and multiple light and dark feature regions in the images related to the process are chosen. Then the spectral curves of these regions are tested, analyzed and summarized. And the maximum spectral differences are compared to judge the spectral consistency. The maximum spectral difference of the two materials are 1.21 μm and 0.30 μm. The results show that the spectral consistency of the optimized process is better than that of the traditional process.