No-reference image quality assessment (NR-IQA) is a fundamental yet challenging task in low-level computer vision. It is to predict the perceptual quality of an image with unknown distortion. Its difficulty is particularly pronounced as the corresponding reference for assessment is typically absent. Various mechanisms to extract features ranging from natural scene statistics to deep features have been leveraged to boost the NR-IQA performance. However, these methods treat images of different degradations the same and the representations of distortions are under-exploited. Furthermore, identifying the distortion type should be an important part for NR-IQA, which is rarely addressed in the previous methods. In this work, we propose the domain-aware no-reference image quality assessment (DA-NR-IQA), which for the first time exploits and disentangles the distinct representation of different degradations to access image quality. Benefiting from the design of domain-aware architecture, our method can simultaneously identify the distortion type of an image. With both the by-product distortion type and quality score determined, the distortion in an image can be better characterized and the image quality can be more precisely assessed. Extensive experiments show that the proposed DA-NR-IQA performs better than almost all the other state-of-the-art methods.