Composite state convergence is a novel scheme applied for the bilateral control of a telerobotic system. The scheme offers an elegant design procedure and employs only three communication channels to establish synchronization between a single-master and a single-slave robotic system. This paper expands the capability of the composite state convergence scheme to accommodate any number of master and slave systems and proposes a disturbance observer-based composite state convergence architecture where k-master systems can cooperatively control l-slave systems in the presence of uncertainties. A systematic method is presented to compute the control gains while observer gains are determined in a standard way.To validate the proposed architecture, MATLAB simulations are performed on symmetric and asymmetric arrangements of single-degree-of-freedom teleoperation systems. Finally, experimental results are obtained using Quanser's Qube-Servo systems in QUARC/Simulink environment.