In video synthetic aperture radar (SAR) imaging mode, the polar format algorithm (PFA) is more computational effective than the backprojection algorithm (BPA). However, the two-dimensional (2-D) interpolation in PFA greatly affects its computational speed, which is detrimental to the real-time imaging of video SAR. In this paper, a terahertz (THz) video SAR imaging algorithm based on chirp scaling is proposed, which utilizes the small synthetic angular feature of THz SAR and the inherent property of linear frequency modulation. Then, two-step chirp scaling is used to replace the 2-D interpolation in the PFA to obtain a similar focusing effect, but with a faster operation. Point target simulation is used to verify the effectiveness of the proposed method.