Sequential recommendation has been widely used to predict users’ potential preferences by learning their dynamic user interests, for which most previous works focus on capturing item-level dependencies. Despite the great success, they often overlook the stage-level interest dependencies. In real-world scenarios, user interests tend to be staged, e.g., following an item purchase, a user’s interests may undergo a transition into the subsequent phase. And there are intricate dependencies across different stages. Meanwhile, users’ behaviors are usually heterogeneous, including auxiliary behaviors (e.g., examinations) and target behaviors (e.g., purchases), which imply more fine-grained user interests. However, existing works have limitations in explicitly modeling the relationships between auxiliary behaviors and target behaviors. To address the above issues, we propose a novel framework, i.e., dynamic stage-aware user interest learning (DSUIL), for heterogeneous sequential recommendation, which is the first solution to model user interests in a cross-stage manner. Specifically, our DSUIL consists of three modules: a dynamic graph convolution module that dynamically learns item representations in each stage, a behavior-aware subgraph representation learning module that learns heterogeneous dependencies between behaviors and aggregates item representations to represent the user interests for each stage, and a sequence decoder to capture the evolving pattern of user interests and make item prediction. Extensive experimental results on two public datasets show that our DSUIL performs significantly better than the state-of-the-art methods.