Parts manufactured by wire arc additive manufacturing (WAAM) are characterized by a peculiar surface morphology, namely, surface waviness, that negatively affects the fatigue performance. To exploit the full potential of WAAM and minimize the need for postproduction work, it is crucial to utilize the components in the as‐built state. This is because conventional machining techniques, typically employed for postprocessing operations, severely curtail the freedom of geometry of the components. This study focuses on an experimental and numerical characterization of the notch effect of the surface waviness for an AISI 308 LSi stainless steel. This is done by quantifying the fatigue notch factor in a probabilistic fashion, considering the results of ad‐hoc designed fatigue tests. A finite element model is developed by considering a 3D scan of the geometry of WAAMed plates, allowing to determine the theoretical stress concentration factor. The fatigue notch factor is also estimated from the numerical model by making use of the gradient correction according to the FKM guidelines. To validate the numerical approach, test data produced for different testing conditions are correlated by using the local stress approach, showing that classical methods are also applicable to additive manufactured parts.