[en] Highlights: • Qualification and assessment of 3D printed components is an open issue • Statistics of extremes applied to CT scans correctly estimates the maximum defect size and the origin of fatigue failures • A simple determination of the fatigue limit and its scatter is obtained • The robustness of the results for the batch investigated is confirmed analyzing the influence of the main parameters involved • Strategies for reducing the quality control time and effort are proposed The progressive improvement of additive manufacturing (AM) techniques enables the production of geometrically complex and lightweight parts, which explains the incredibly fast growth of AM for space and aerospace applications. The first standards require the production of witness specimens manufactured together with the components for assessment and qualification for the material control. However, one of the most critical issues is determining the ‘fitness-for-purpose’ of fatigue loaded parts, which is strictly related to the microstructure and defects generated during the manufacturing process. One of the most suitable techniques for detecting defects even near the surface or in thin, complex geometries is micro-computed tomography (μ-CT). This paper deals with the application of statistics of extremes for analysing X-ray CT scan measurements in view of component assessment by discussing its advantages and requirements.