Abstract: Graph Convolution Networks (GCNs) are a powerful approach for the task of node classification, in which GCNs are trained by minimizing the loss over the final-layer predictions. However, a limitation of this training scheme is that it enforces every node to be classified from the fixed and unified size of receptive fields, which may not be optimal. We propose ProSup (Progressive Supervision), that improves the effectiveness of GCNs by training them in a different way. ProSup supervises all layers progressively to guide their representations towards the characteristics we desire. In addition, we propose a novel technique to reweight the node-wise losses, so as to guide GCNs to pay more attention to the nodes that are hard to classify. The hardness is evaluated progressively following the direction of information flows. Finally, ProSup fuses the rich hierarchical activations from multiple scales to form the final prediction in an adaptive and learnable way. We show that ProSup is effective to enhance the popular GCNs and help them to achieve superior performance on miscellaneous graphs.