The epistemologically distant nature of many of today's environmental risks greatly problematises conventional risk analyses that emphasise objectivity, materiality, factual specificity and certainty. Such analyses fail to problematise issues of ontology and epistemology, assuming a reality that is readily 'readable' and a corresponding knowledge of that reality that is asocial, objective and certain. Under the weight of modern, invisible, manufactured environmental risks, however, these assumptions begin to crack, revealing their tenuous nature. As this paper argues, statements of risk are ultimately (...) social products that come to us by way of translation. They are statements not of what is but of knowledge expressed in probabilistic terms, and are thus thoroughly social in nature, for it is we – through our actions and social networks – that imbue them with meaning. One way we do this is through our social relations of trust. And it is this relationship – between trust and risk – that this paper seeks to detail both conceptually and empirically. While one could look toward any number of case studies to develop the conceptual details of this project, this paper focuses on the relatively recent case of mad cow disease to have been reported within the United States. Here, we have an epistemologically distant, and thus hotly contested, 'object', which has been the source of much risk debate; a debate that is also, in part, the effect of a deeper erosion of trust, particularly toward those managing our meat supply. (shrink)