In this paper, an event-based leader-following strategy for synchronization of multi-agent systems (MASS) is considered. A model-based approach is adopted to predict the relative inter-node states between intermittent communications. The predicted values of the relative inter-node states are forwarded to the controller to calculate a piecewise continuous control signal. The communication between two linked agents is triggered according to a protocol based on their control inputs. The proposed leader-following strategy guarantees exponential state synchronization under time-dependent thresholds and bounded state synchronization under constant thresholds, respectively. Furthermore, the elapsed time between any two successive triggering instants for any pair of linked agents is lower bounded by a constant. The communication frequency reduction potential of the proposed leader-following strategy is well demonstrated via a numerical example. (C) 2016 Elsevier Ltd. All rights reserved.