Obtaining proficiency in interpreting has traditionally rested on two pillars: class sessions, where students receive teacher input on their performances, and private study sessions, where they rely on self-assessment and peer evaluation. As a rule, and certainly if there is no proper guidance, both centre heavily on the final product – the interpreted target text – and on repetition of the interpreting process as a whole. Drawing on the theories of capacity allocation, deliberate learning, and automation, this article proposes to complement these two pillars with a third: that of skill-oriented rather than product-based self-study.