Two kinds of testing devices, TR-I and TR-II, were developed in order to identify varietal differences of rice in shattering habit in a quantitative term. Forty-two varieties were tested with these equipment. There were close relationships between the quantitative measurements taken by means of these two devices and also between the processing losses of a combine harvester and the threshing rates with TR-I as well as the straight tensions with TR-II. The proposed method with those devices provides a quantitative basis for evaluating a degree of shattering of rice varieties without field testing and predicting losses in processing with a combine harvester.