The industrial material test must be simple. In this sense, COD bend test is thought to be an effective testing method. But one of the troubles in the conventional COD bend test is that the clip gauge readings (Vg) must be converted to the notch tip COD by means of rotational factor γ which is to be experimentally determined as a function of V _g or the fracture load for a given material. The authors proposed in the previous paper that the critical COD could be evaluated simply from fracture load and yield stress with the use of the relation of E Φ/ d σ _Y vs. σ _N /σ _Y ( E : Young's modulus, Φ : notch tip COD value, d : specimen depth, σ _N : nominal skin stress at the notched section, σ _Y : yield stress) obtained from FEM. In this paper, it is newly proposed that the method using FEM is to be applied with the limitation of σ _N /σ _Y _??_1. 5. For the range of σ _N /σ _Y _??_2. 0, the rotational factor γ=1/3 is to be used taking advantage of the constancy of γ when considerable deformation occurs prior to fracture. Thus the experimental determination of γ is not needed. For the intermediate range of 1. 5_??_σ _N /σ _Y _??_2. 0, the above two methods are to be combined, i. e. Φ _c for bend test may be estimated by a fair curve connecting the points of Φ _c obtained by using FEM at σ _N /σ _Y =1. 5 and Φ _c obtained by using rotational factor at σ _N /σ _Y =2. 0. (Φ _c : critical COD value) The proposed standard specimen size is identical to ASTM recommendation. It was confirmed that COD bend test is one of the most effective and useful test methods from the fact that critical COD obtained from bend test compares very well with that obtained from tension test.