In these 10 simple rules for building and maintaining responsible data science workflows, we walk through the lifecycle of a project and consider how a research team can responsibly manage both the technical and social aspects of the project, adapting when necessary. These 10 rules are by no means prescriptive as we recognize the complexities surrounding responsible research and the heterogeneity of data science applications across research communities and fields. In addition, the iterative nature of exploration and refinement within a project can lead to nonlinearity in the workflow that can make data and computationally intensive research challenging. Nevertheless, we hope the rules can help interested researchers build and maintain a responsible workflow and collaborations.