In this article, a robust consensus tracking problem for a multi-agent system with integrator-type dynamics has been addressed in the presence of disturbances and unmodelled dynamics. The desired trajectory to be tracked is only provided to a small group of team members. An identifier is developed to estimate the unknown disturbances and unmodelled dynamics. A consensus controller is developed based on this identifier to achieve asymptotic consensus tracking using the local information obtained from neighbouring agents. The closed-loop stability is proved using the Lyapunov-based analysis and an invariance-like theorem. Numerical simulations are provided to demonstrate the effectiveness of the developed robust consensus controller.