Abstract: The ordinary MRTD-CW model has been used to evaluate the comprehensive performance of thermal imaging systems, but can't be used to evaluate the influence of natural background on system performance. In contrast to the targeting task performance model (TTP), a modified MRTD-CW model was proposed by introducing the natural background parameters to reflect the influence of nature background on comprehensive performance of thermal imaging systems. The natural background parameters were calculated by noise standard deviation based on typical scenes viewing experiments as example. MRTD-CW model with the natural background parameters was used to calculate the target comprehensive information quantity acquired by thermal imaging systems at different scenes viewing experiments, and the optimal angle magnification calculated by the modified MRTD-CW model was very close to that of TTP. The experimental results show that MRTD-CW model with the natural background parameters can analyse effectively the influence of nature background on comprehensive performance of thermal imaging systems.