The recent rise of computation-based methods in social science has opened new opportunities for exploring qualitative questions through analysis of large amounts of text. This article uses a mixed-methods design that incorporates machine reading, network analysis, semantic analysis, and qualitative analysis of 414 highly cited publications on obesogenic environments between 2001 and 2015. The method produces an elaborate network map exhibiting five distinct notions of environment, all of which are currently active in the field of obesity research. The five notions are institutional, built, food, family, and bodily environments. The network map is proposed as a navigational tool both for policy actors who wish to coordinate efforts between a variety of stakeholders and for researchers who wish to understand their own research and research plans in light of different positions in the field. The final part of the article explores how the network map may also initiate a broader set of reflections on the configuration, differentiation, and coherence of the field of obesity research.