Several testing methods have been proposed for evaluating the mode I fracture toughness of rocks. Since the fracture toughness evaluated from the different testing methods is generally not compared, what is still not wellknown is the effect of the testing method type on the fracture toughness. The effect of not only the testing method type but also the specimen size exists in the fracture toughness of rock materials. To clarify the effect of the testing method type and specimen size on the fracture toughness, many fracture toughness tests were performed in the previous experimental studies. However, the effect has not been elucidated because the number of the specimens is limited and the range of the specimen size is relatively small in the previous works. In this paper, firstly, three types of the fracture toughness test, the chevron bend (CB) test, semi-circular bend (SCB) test and straight notched disk bending (SNDB) test, were performed using Kimachi sandstone. Then, the SCB test was carried out using the same sandstone with various specimen sizes. As a result, the fracture toughness obtained from the CB and SCB tests were compatible each other. However, the SNDB tests estimated lower value of the fracture toughness than the others. From the results of the SCB tests with various specimen sizes, the fracture toughness increased with increasing radius. Based on the results, the effect of the specimen size on the fracture toughness and the differences of the fracture toughness obtained from the three testing methods were discussed.