In order to arrest the propagation of brittle crack in welded steel structure, various means (for example, rivet seam, insertion of very notch tough strip, etc.) have been devised and adopted. Up to this time, however, only a few analytical or experimental researches are available concerning those crack arresters. It may be no exaggeration to say that these are no data which can be usable for practical crack arrest design, for example, to determine most suitable diameter and/or pitch of rivet under given design data (service stress, lowest service temperature, plate thickness etc.). The present authors have been engaged in the researches on brittle fracture of structural steels from the standpoint of “Fracture Mechanics”. In this report the authors tried to apply “Fracture Mechanics” to analize the role of two simple and fundamental crack arrester models (1. Single or double circular holes located near the pressumed crack propagation line. 2. Rivet seam) and carried out some experiments using “Double Tension Test” specimen. As a result of analysis on experimented data, it has been made clear that brittle crack is to be arrested when strain energy release rate or K-value (Irwin's stress intensity factor) of a crack running through the stress field with arrester model falls down below the Kc-value at that temperature obtained from Double Tension Test. In other words crack arrester can display its effective role when it can make a minimum in K-curve and the minimum value is smaller than that abtained from Double Tension Test at that temperature. It can hardly be possible to draw any definite conclusions, because of insufficiency of specimens tested to verify or compare to the results from analytical calculation of K-value.