In this paper, we address the leader-following consensus problem of linear multi-agent system with improved transient performance. A state feedback controller incorporating a novel reset mechanism is proposed. It is shown that the convergence rate of the closed-loop system can be improved with the introduction of the reset mechanism. Furthermore, the regularity of the reset mechanism can be ensured for all agents, in contrast to previous studies where it was guaranteed only for one agent, which implies that the reset controller is bound to be triggered in finite time. Finally, an example is provided to verify the advantages of the proposed controller over some other kinds of control methods.