Abstract: After more than ten years of development, great achievements of soft-sensing technology have been made not only in theory but also in practice.Up until now, however, most research work in the soft-sensing field have been essentially concentrated on static soft-sensing models.The relations and distinctions between dynamic soft-sensing models and static soft-sensing models are discussed. Using the impulse response model, a dynamic soft-sensing method based on least squares is developed. Some improvements to the method are also presented. A simulation example is given to validate the effectiveness of the method and its improvements.