Quality control is the key issue that needs to be addressed in any gas metal arc welding (GMAW) system, especially in robotic pipeline welding system. This paper explores a second-order sliding mode control (SMC) strategy—a variable gain super-twisting control, to maximize the productivity, consistency in welding quality. This is achieved by the robust finite time output tracking of GMAW system. A nonlinear multi-input multi-output (MIMO) model of GMAW system has been considered here for the design of variable gain super-twisting (VGST) controller by which complete rejection of the bounded uncertainties/disturbances is possible and the adaptive characteristic of its gains help to use the control effort effectively. The stability of internal dynamics of the system is studied to establish the feasibility of solving the robust finite time output tracking problem. The stability of the overall system has been analyzed using Lyapunov stability criterion. The performance of the controller is demonstrated using the model of the system emulating the realistic conditions of operation. The simulation results are presented to illustrate the efficacy of the controller.