Ultrasonic testing is the most common non-destructive testing method to detect flaws in large heavy forgings such as turbine rotors. Flaw sizing methods as well as theoretical calibration methods apply artificial defects such as flat bottom holes. While some calibration methods have been verified for certain UT applications and materials, others still need verification. In order to do so, artificial defects of high quality are required. This report deals with an EDM process to produce flat bottom holes of approximately 2 mm diameter and high aspect ratio of about 5. 16 FBHs were produced with varied parameters and measured by destructive and non-destructive testing methods. The aim was to identify a suitable NDT technology measuring FBH quality as well as most suitable parameters providing best FBH results. Impacts of EDM process details on flat bottom hole results were analysed, affecting ultrasonic testing quality. Particularly tool wear and flushing conditions were identified to have a major impact on geometrical properties. For ultrasonic testing at 2 to 5 MHz, flatness is less important than corner curvature due to electrode corner wear. This is explained by UT detecting only a material height of 1000 μm or 333 μm respectively. Comparably low pulse frequency with 20 μs on-time proved to deliver most suitable results. Regarding FBH quality examination, plastic replication using a two-part silicon achieved sufficiently precise results.