Graph Neural Networks (GNNs) have recently received significant research attention due to their prominent performance on a variety of graph-related learning tasks. Most of the existing works focus on either static or dynamic graph settings, addressing a particular task, e.g., node/graph classification, link prediction. In this work, we investigate the question: can GNNs be applied to continuously learning a sequence of tasks? Towards that, we explore the Continual Graph Learning (CGL) paradigm and we present the Experience Replay based framework ER-GNN for CGL to address the catastrophic forgetting problem in existing GNNs. ER-GNN stores knowledge from previous tasks as experiences and replays them when learning new tasks to mitigate the forgetting issue. We propose three experience node selection strategies: mean of features, coverage maximization and influence maximization, to guide the process of selecting experience nodes. Extensive experiments on three benchmark datasets demonstrate the effectiveness of ER-GNN and shed light on the incremental (non-Euclidean) graph structure learning.