Welded joints are frequently used in machine construction, ships, and bridges and must function safely throughout their service life as determined for certain circumstances defined in design codes, standards, and guidelines. Although complex analysis of welded joints gives satisfactory results, factors that modify the endurance limit are common in industrial use. The endurance limit-modifying factors given in reference books are unable to generalize to all types of welded joints. Furthermore, cracks in fillet welded T-joints are likely to start from the toe of the weld bead; however, the weld throat thickness is considered in the calculation of the strength of a welded joint. In this study, we examined the stress concentrations of load-carrying welded T-joints considering bead shape and thickness via finite element analysis and verified the model experimentally. We utilized an electromechanical cylinder to carry out experiments to obtain stresses near the weld bead via strain gauges. The submodeling technique was implemented to obtain results in the regions of stress concentrations as accurately as possible. The results of finite element analysis were in good agreement with experimental results. The present study showed that the ratio of weld throat thickness to plate thickness significantly affects the stress concentration factors of load-carrying welded joints. The maximum stress decreased significantly depending on bead shape and thickness. Endurance limit-modifying factors gathered via analyses assuming the weld as a notch and considering plate thickness could be used in the fatigue strength calculations of welded joints.