The first stages of the modeling process remain thus far poorly studied, in particular the modeler’s decision-making in regard to the inclusion and representation of environmental phenomena within the model. Anchoring itself in the theoretical and methodological framework of Science and Technology Studies, the present study aims at identifying and retracing the origin of the modeler’s choices. Extensive, semi-constructed interviews were conducted with European and North American modelers in a variety of disciplines within Earth and Universe sciences. The outcome of the interviews sheds light on the role of different actors in transferring choices progressively incorporated in the modelers’ practice, up to becoming self-evident. We introduce and engage the sociological concept of habits as an additional layer of analysis. Modeling habits can contribute to efficiency and expertise – provided their origin, epistemic load and implications in regard to collectives are assessed.