This article moves into the complex relationship between law and (scientific) expertise. The author firstly discusses the difference between scientific knowledge and scientific expertise, tracing the historical roots of the concept of proof in a legal rather than a mathematical context. Secondly, the historical roots of the fair trial are discussed in terms of the pre-modern épreuve and the modern preuve, demonstrating the relationship with coordinate and subordinate types of justice. Thirdly, Foucault's analysis of épreuve, enquête and examèn is extended to clarify how preuve and épreuve have been integrated into the fair trial, which is explained in relation to the formal and the substantive notions of the 'Rechtsstaat'. This analysis finally allows the author to discuss the difference between two approaches of scientific expertise in court: one claiming that judges should defer to science and another claiming that a court of law is one of the spaces in which such expertise can be contested. The article concludes that the integration of subordinate and coordinate justice exemplified in the fair trial, opens new perspectives for the testing of scientific expertise, taking into account whose interests are at stake.