The paper deals with the finite horizon H_infin white noise smoothing problem for a linear time-varying discrete-time system. A new approach termed as re-organized innovation sequence is presented to derive the fixed-lag smoother. With the new approach, it is clearly shown that the central white noise smoother in an H_infin setting is the same as that in an H₂ setting associated with one self-constructed stochastic state-space model. The solution to the H _infin white noise smoother is given in terms of solutions to Riccati equations with the same order as the original system