Many kinds of weld cracking test such as slit type restraint cracking tests and tensile restraint cracking tests have been developed for assessing of weld cracking susceptibility of high tensile strength steels and their weld metals. However, there have been few papers which are dealing with the correlation among the results obtained from these different kind of testing methods, and also the cracking encountered in the actual structures. Consequently, the investigation of the effects of shapes and sizes of test specimens on the cracking will contribute to the solving of the above-mentioned problems. From this point of view, the authors carried out the y slit type test of so called Tekken type and tensile restraint cracking test by using of 80 kg/mm² high tensile strength steel of 50 mm in thickness. In tensile restraint cracking test, the specimen were made with the variation of 120 mm1, 000 mm in width, and studied the effect of residual stress and cooling time on the cracking tendencies in those tests. From the results of these experiments the following facts are clarified; (1) In the tensile restraint cracking tests, the critical restraint stresses are recognized to decrease to some extents the width of specimens increase. This fact might be attributed to the lower level of residual stress existing in narrower specimens. (2) In comparison of test results of y slit type and tensile restraint cracking tests, y slit type tests are likely to show a little lower cracking susceptibility. This is due to the difference of cooling time and so both y slit type test and tensile restraint tests indicate the fairly good correlation of testing results when they are shown in relation between the restraint stress and the cooling time down to 50°C. (3) The results of y slit type tests can be applied to the decision of preheating temperature of actual structures if their joints would be preheated uniformly or the post-heating as well as preheating would be done.