The nationalization of Iran’s oil industry in 1951 is often held to be a key turning point in the story of Middle Eastern oil, supposedly marking when the domination of transnational oil corporations started to wane amidst popular nationalism’s demand for sovereignty over natural resources. But in this focus on sovereignty, there has been a tendency to take for granted the conversion of a subsoil resource into national revenue, overlooking the ways in which oil is produced through a set of arrangements intricately linked with politics, as Timothy Mitchell argues. The nationalization of Iranian oil has been studied little with a view to the particular assemblage of knowledge, practices, technologies and infrastructures that produced oil in the first place. This paper turns attention to this oil assemblage by examining the Anglo-Iranian Oil Company’s operations in the Abadan refinery in c.1939-51. Refineries were key nodes for the standardisation of oil according to the transnational regulatory norms of what Barry terms a ‘technological zone’, helping companies in their attempts to create a seamless, controlled flow of oil from the wellhead to the individual consumer. In Abadan, aside from crude oil itself, one overlooked local factor that created friction to flow was labor. The subjectivities of this ‘human factor’ had to be calculated, controlled and augmented through disciplinary mechanisms in the same way as crude. Through these processes, expertise was produced through the separation of mind and body, whereby materialities (including workers’ bodies) were subjugated to abstract, disembodied knowledge in refinery operations. The refinery, then, was supposedly an automated, self-regulating and objective “city of science” disentangled from locality, including politics. This helped conceal the highly exclusionary practices that (re)produced expertise as a white, male domain. Using archival sources from Iran, the UK and the US, in addition to newspapers, oral histories and memoirs, in this paper I highlight both how this production of expertise worked in practice and how it was contested. I show that workers challenged it in two ways: first, by resisting disentanglement through acts of ‘everyday resistance’, sabotage and strikes, connecting the refinery to the politics of the town; second, by undermining the very ontological basis of this expertise through their subjectivities and ‘tacit knowledge’ blurring mind-body dualisms. Ultimately, I argue, this contributed to nationalization’s demand for access to expertise. However, nationalisation did not contest the content of that (seemingly objective) expertise, helping transnational oil corporations in the years ahead.