In this paper, I consider recent work which has drawn on the wealth of material in the Stanley Milgram Papers archive at Yale University to generate new insights into Milgram’s ‘obedience’ experiments. I will suggest that this work has begun to reshape how we understand these experiments, and in particular will draw attention to the way in which audio recordings of the experiments point to the role of rhetoric in the experiments, and highlight hitherto unacknowledged procedural flexibilities in the standardized experimental protocols. These points are illustrated using a revised transcription of one of Milgram’s case examples to further highlight the extent to which Milgram appears to have ‘smoothed over’ his transcript. But I will also suggest that the capacity for qualitative analysis to be enriched by engagement with archival material should also ensure that the insights of poststructuralist and constructionist approaches are brought to bear on the archives themselves. Drawing on Derrida’s Archive Fever, and on the literature on transcription practices in qualitative research, it is argued that as a source of data, we should apply the same cautious skepticism to archival material as we would with any other data. When faced with a resource as rich and compelling as the Stanley Milgram Papers, the temptation may be to reify the materials held there as offering new and unvarnished ‘truths’ about what ‘really’ happened in the experiments. By contrast, it will be suggested that as well as a resource, archives should be a topic for analysis.