This article aims to predict the crack-initiating defects and to estimate the fatigue limit of additively manufactured Ti-6Al-4V specimens with Murakami’s empirical √area-parameter model. For this purpose, fatigue specimens are manufactured by laser powder bed fusion (LPBF) using different process parameters and selected heat treatment methods. The specimens are scanned using xCT in two different manners: first full in reduced resolution to scan the complete geometry and second the ROI in high resolution. Afterwards, the specimens are tested cyclically in the very high cycle fatigue (VHCF) regime. Crack-initiating defects are analyzed post-mortem using scanning electron microscopy (SEM). The defect population is examined in the xCT data, in which crack-initiating defects are identified. Using the empirical √area-parameter model, the effect of defects on the potential fatigue location is determined. An outlook is given on the performance of the empirical √area-parameter model on the prediction of crack-initiating defects and fatigue strength of additively manufactured Ti-6Al-4V specimens.