The proposed research project addresses the visualization of air pollution in Southern California, driven by the larger question of why it is so difficult to ‘see toxics’. For this purpose, the ethnographic fieldwork will attend to already existing visualization tools, their genealogies, and the identification of concerned actors. Starting points for collaboration will be interviews with the Office of Environmental Health Hazard Assessment (OEHHA), which developed the ‘CalEnviroScreen 3.0’,data mining research groups like ‘CrossTown’ at USC, the visualization team at the Los Angeles Times and local environmental justice groups. There, the analytical focus will be on the limits of imagining, infrastructuring, and visualizing toxicity in contemporary data-driven practices. Such limits might be found in the structural absence of data, their contested preservation, and inadequate interpretive capacities. Together, these might give form to what I will provisionally call ‘toxic expertise’. In this light, the study aims to help better understand present ‘data-to-governance’ pathways in Southern California, including knowledge infrastructures threatened by public funding and other unarticulated yet crosscutting inequalities. Overall, the findings aim to lend itself for comparative analysis on the city or federal level, as carried out in collaborative research endeavours like the Asthma Files.