Abstract: Laser selective melting (SLM) serves as a pivotal technology in the additive manufacturing of aero-engine components. To address the issue of micro-defects, a micro-CT detection method is proposed. A complex-structured fuel nozzle is used as the research subject. Firstly, comparative samples with artificial defects are manufactured. The optimal parameters for CT detection are obtained through tomography experiments. Afterwards, taking the fuel nozzle characterizaed with natural defects, SLM-shaping, and complexed construction as target, detection on micro-focus CT defects was conducted using the optimal dectection parameters that have been achieved. The results indicated that pore-like and point porosity defects can be clearly displayed using the same parameters as those for the sample. The minimum detectable size of defects is 95 μm, and the process sensitivity requirement for detecting 100 μm defects is fulfilled. The proposed method provides a new technological approach for improving product quality.