My research project will focus on analyzing the role of ‘levels of organization’ in scientific reasoning about explanation in biology. The concept of ‘levels’ evokes a hierarchical image of the world that is vertically stratified into a graduated continuity: The things found at one horizontal slice of the world somehow 'make up' or 'are continuous with' things found at another slice, and so on. Yet, despite its ubiquity, ‘levels’ remains a decisively ambiguous concept in biology. The 'things' that constitute the horizontal image of the world can vary widely between different contexts. Sometimes they refer to real objects, like part-whole compositional relationships, or they can refer to epistemic units, such as the “level of genetics” or “level of ecology”. In my dissertation, I showed that this variability reveals usage of ‘levels’ in science to be governed mostly by its intuitive appeal, whose justification is based on context-dependent criteria of adequacy. Far from useless, however, ‘levels’ instead exemplifies what one philosopher calls “productive ambiguity” of open-ended helping concepts in science. For this reason, philosophical analyses should abandon attempts to unify different uses of ‘levels’ and acknowledge the fragmentary character of the concept. This research project will expand on this foundation by explicating two distinct roles attributed to the concept of ‘levels’ in different arenas of scientific reasoning concerning explanation. One of these roles is pedagogical, and is often encountered in introductory textbooks to biology, such as the Campbell Biology series, where ‘levels’ is used to introduce the major unifying themes of the field. Another role is organizational, and is encountered in professional research literature of working scientists, in particular review articles and commentaries on the state of research in a certain area of investigation. Here ‘levels’ is often used by scientists as a theoretical device with which to navigate within complex phenomena distributed across multiple levels, or to coordinate interdisciplinary efforts for investigating such phenomena that cannot be adequately addressed by any one of the involved disciplines alone. Though distinct, these roles complement one another in biological reasoning, albeit in starkly different contexts of scientific practice. To conduct this analysis I will focus on two case studies from science in which levels are prominently used in both of the roles specified above. The first, macromolecular structure, is treated in a well-defined local manner, focusing on features of (some) macromolecules such as nucleic acid and proteins (but, tellingly, not lipids or sugars). The second, the nervous system, is treated in a more open-ended general manner, owing to the vast complexity of neural phenomena that ‘levels’ is used to investigate.