This paper is concerned with finite-time L₂ leader-follower consensus of networked Euler-Lagrange systems in the presence of external disturbances. A distributed finite-time L₂ control protocol is proposed by using backstepping design such that a group of follower agents modeled by Euler-Lagrange systems can follow a desired leader agent and achieve leader-follower consensus in finite time. Moreover, the finite-time L₂ gain is less than or equal to a prescribed value. A simulation example of a network composed of seven two-link manipulators is given to show the effectiveness of the theoretical results.