The precracking requirements for fracture toughness testing of metallic materials are investigated in this study. In addition, some experiments were performed in the transition region to quantify the consequences of violating these requirements. It is found that the different standards are not consistent. For most of the requirements some rationale can be found. Some theoretical developments were performed to justify the use of a fixed maximum stress intensity factor. This limit is tightly related to the fracture mechanism and to the fracture process zone. From an experimental point of view, a relatively limited effect of pracracking on the fracture toughness is found.