The spectral characteristics of sensor sensitivities are an important component of image signal formation and also essential knowledge for illuminant estimation methods. Due to the wide variation of sensor spectral sensitivities in cameras, the efficiency of illuminant estimation methods could also be affected. In this paper, we study the influence of sensor variation on the efficiency of different illuminant estimation methods, and find out the type of sensor sensitivities that have better performance for corresponding methods. Also, we propose a way of using spectral sensitivity replacement for methods to be applicable with unknown sensor information. The optimal replacing sensor sensitivities can be obtained through testing on a series of sensor combinations. In addition, the testing result reveals the efficiency descent degrees for those methods when using the replacing sensors. Those insensitive to incorrect sensor information are better methods to be applied to images with unknown spectral sensitivities.