This paper presents a new networked hierarchical predictive sliding mode controller for the design of stabilizing signals to damp low frequency oscillations observed in geographically distributed networked power systems subject to communication network constraints. Firstly, through the generic framework of networked systems, a networked model is used for a distributed networked system in which network delays and data packet loss are considered in a wide area measurement system (WAMS). Then a hierarchical predictive sliding mode controller with robust reaching law is proposed to compensate the influence of WAMS communication latency and damp the inter-area oscillations. For this, firstly, the overall system is decomposed into small subsystems with non-sparse matrices and lower order; then a new suitable predictive sliding mode controller is designed to provide optimal performance. Furthermore, in order to coordinate the entire system, the gradient of interaction errors is used. The effectiveness of the proposed control framework is illustrated through numerical simulations on a 5-area, 16-machine power system with several interconnected areas. Simulation results show the capability of the proposed approach to enhance the networked power systems' damping in comparison with a conventional networked control method. KEYWORDS hierarchical control, networked control, power system stability, predictive sliding mode, wide area system 1 | INTRODUCTION Distributed wide-area systems commonly appear in the real world, in the form of, for example, power grids, chemical plants, water distribution networks, traffic systems, etc. Systems are wide-area systems if they can be decoupled or decomposed into a number of interconnected/interacting subsystems or if their dimensions are so large that using the usual analysis, optimization, and control methods cannot achieve an acceptable answer during reasonable computational time and efforts [1]. Such typical motivating problems arise in the control of widely distributed multi-area power systems with severe interactions. These problems are mainly due to large dimensions, modeling errors, disturbances, and uncertainties in such systems. The stability of modern power grids is one of the most important issues in the operation of these systems. The increasing size of wide-area electric power grids and the interconnection of these systems via weak tie