<p>Wire arc additive manufacturing (WAAM) is a promising additive manufacturing (AM) technique for fabricating large net shape parts. However, to confidently use such parts in structural applications, part qualification with respect to fatigue is required. This study on the effect of defects on S-N performance using established qualification procedures reveals new insight into reference sample and part-extracted sample equivalence, fatigue sample size effect, applicability of defect tolerance models, and crack initiation life prediction from X-ray CT detected defects. In this study, in-process qualification of the less commonly reported plasma arc welding technique ER70S-6 parameters for fatigue was performed using specimens extracted from a wall for tensile, S-N, and fatigue crack growth rate (FCGR) tests. Part qualification was then performed for an actual full-size pressure vessel using specimens cut from it for S-N tests. In the process, a methodology was established for part qualification using ASTM standards. Pre-testing X-ray computed tomography (CT) was used in conjunction with fractography to gain insight into the influence of defects on fatigue performance. Naturally occurring large defects provided valuable data for studying crack growth around them. The S-N curves from both sets were very similar. In the last section, recommendations for predictive techniques are presented. X-ray CT data processing for crack initiation life prediction showed that crack initiation occurred in a small portion of the total fatigue life for a large spherical defect. Defect tolerance models were adapted, and for this alloy, the Kitagawa-Takahashi model described the experimental data more closely than the El Haddad and Murakami empirical models.</p>