Pesticide use and exposure have reached a truly planetary scale. Pesticide dependence is a defining feature of modern agrarian capitalism, and is growing rapidly in low-income countries. Yet the debates on pesticides, development and health have continued on broadly similar terms for more than 50 years. In this dissertation, I outline the changes in the global political economy of pesticide use in the last twenty years, including changes in the production of off-patent pesticides, and drivers of increased adoption rooted in longstanding processes of agrarian change including expanding global commodity markets, rural-urban migration, and rising consumption in the global south. Understanding what this shift mean, and evaluating its consequences, requires thinking differently. I use a case study in Lao PDR, until recently one of the world’s lowest per-capita pesticide users, to try to chart a different path through these debates, interrogating narratives of development, risk and uncertainty. In a part of the country where maize and herbicides have completely transformed the countryside, I use oral histories, interviews, surveys and agroecology to look at the relationship between modern agriculture, intensification, precarity, and deforestation as old forest-based safety nets were ploughed under for monoculture maize production. I interrogate the diverse -- and divergent -- set of partial knowledges among pesticide users, and ask how pesticide applicators’ direct experiences of toxicity both reinforce and transgress the dominant model for ‘safe use’ of pesticides. Drawing on contextual understandings of risk from feminist science studies and political ecology, I understand how users grapple with risk and uncertainty in ways that produce patterns of unequal exposure on the local level, that produce a class of ‘risky subjects’. I also follow a community-based biomonitoring program and follow the controversy over the data they produced. I use this controversy to illuminate a larger one. Debates about what counts as harm, what evidence is valid, and what science is adequate proof have a history almost as old as chemical agriculture itself. The domain of imperceptibility surrounding exposures, the geographic disparities in terms of the ability to make impacts visible, and the disciplinary ‘tracks’ that limit what kind of evidence count, make it very difficult to produce enough scientific certainty to know the magnitude of the problem.