Abstract: An improved finite-difference time-domain (FDTD) method with the semi-analytical recursive convolution (SARC) algorithm in digital signal processing (DSP) techniques for analysis of optical characteristics of the dispersive metals is presented. In this method, the absolute stability, high accuracy and low storage of the original SARC FDTD in dealing with general dispersive media is retained, while a concise update formula with increasing computational efficiency is presented by introducing the trapezoidal approximation. The proposed SARC FDTD approach can therefore be applied to the analysis of optical characteristics of metals provided the pole and corresponding coefficients in the metal dispersion model of interest are given. Three typical kinds of dispersive model for the metals, i.e. multi-poles Lorentz, Drude-Lorentz and Drude-CP are tested to demonstrate the feasibility of the present scheme.