Torture has a complex relationship with technology. Technology variably facilitates or hinders modes of inflicting and investigating torture. Torture itself remains a technology for producing proof, simultaneously primitive and sophisticated. Torture techniques and practices evolve and travel, are prevalent in both democratic and autocratic states, and are often accompanied by a pseudo-scientific rhetoric of ‘controlled’ pain. Conversely, juridical processes pursuing accountability for torture also focus their evaluative gaze on the tortured body and to scientific expertise. To this end, sustained transnational efforts have culminated in the crafting of evidentiary technologies to document torture as grounded in state-of-the-art medical and psychological knowledge. What is more, despite this demand and supply, the utility of this ‘technology of knowledge’ remains unfulfilled in practice for different reasons. I start with the dialectic interplay between infliction and investigation produced by these practices—with prevalent methods informing documentation practices, and with those responsible for infliction adjusting their practices accordingly to evade accountability. With the long-anticipated revised version of the UN Istanbul Protocol now at hand, I revisit the origins and promises undergirding these efforts and their implications for anti-torture practice. I conclude with a critical reflection on the contested dynamics of knowledge imbued in these practices—asking how we (and the law) can ever adequately know about the tortured body.