<p>The Convolutional Neural Networks (CNNs) have achieved outstanding performance in the field of Image Super-Resolution (SR). However, many existing methods focus excessively on increasing network depth, which leads to an excessive number of parameters and computational complexity. Some methods, aimed at reducing model capacity, significantly reduce the receptive field of the network, and employ a single upsampling method, resulting in suboptimal model performance. To better balance model size and performance, this paper proposes a lightweight Adaptive Information Fusion Attention Network (AIFAN). Specifically, a Dual Path Block (DPB) is designed to adaptively aggregate multi-scale features from different paths through the Adaptive Path Fusion Block (APFB). To take full advantage of the hierarchical features, all DPBs are connected using a dense connection mode. Meanwhile, a receptive field attention (RFA) module is proposed to increase the receptive field of the network and force the network to pay attention to the region with more information in the feature. In addition, a hybrid up-sampling module is designed. The Adaptive Multi-scale Up-sampling (AMSU) module and Feature Update Up-sampling Block (FUUB) are combined in the module to obtain better reconstruction results. The experimental results show that the proposed AIFAN achieves an effective balance between performance and model complexity.</p>