This study presents a genetic algorithm-based methodology for reconstructing the nominal profile of airfoils belonging to the NACA four- and five-digit series. By minimizing the geometric deviations between measured point clouds and parametrically generated airfoil profiles, the algorithm identifies the best-fitting nominal geometry. The approach was implemented using Rhino 8, Grasshopper, and the Galapagos plugin, and validated through extensive testing on 3D-printed samples. Across 200 test runs, the algorithm consistently identified the correct nominal geometry, demonstrating robustness despite inherent stochastic variability and computational challenges. The average number of iterations needed to converge was found to be 953 across all cases. This methodology offers a valuable tool for reverse engineering and metrological applications, providing a parametric and efficient alternative to traditional free-form surface reconstruction.