<p>Welded structures are subjected to various repetitive loads during operation. Such repeated loads may initiate fatigue cracks and eventually cause fatigue failure in structures. Fatigue failure prevention is one of the most important issues to be carefully considered in the assessment of the structural integrity. Therefore, numerous studies have been conducted on fatigue performance evaluation of welded structures. During actual operation, various mean stresses act on welded structures, and these mean stresses have a significant influence on fatigue life assessment. However, there is not sufficient research explicitly explaining the mean stress effect in low cycle fatigue regime, especially when compressive mean stress is included. In this regard, the mean stress effect in low cycle fatigue regime was investigated. The target material is high strength steel with yield strength of 600 MPa or above that is commonly used due to excellent mechanical properties and weight reduction. A series of fatigue tests were performed on the high strength welded T-joint under various stress ratio conditions. Conventional mean stress correction models such as Goodman and Soderberg models are attempted for the interpretation of LCF results. However, the conventional methods are not successful to effectively correlate the mean stress effects. Therefore, relatively new methods, structural strain and effective mean stress methods, are applied. The results applying the structural strain method exhibited improved mean stress correction results by gathering the fatigue test data within a narrow band with test data in various mean stress conditions. From the results of this study, it is expected that the mean stress effect can be considered in more consistent manner for fatigue life assessment using structural strain and effective mean stress methods.</p>