161 cases of liver cirrhosis were selected for studying the correlation between their abnormality of glucose tolerance and the pathologic stage or clinical course of them. The cases were divided into two groups; a group of 42 cases were clinically considered to have combinated with primary diabetes and the remaining 119 cases were not. Besides, the latter were classified more small groups according to the results of GTT. These groups of cases with liver cirrhosis were compared with the groups of probable diabetics, chemical diabetics and the clinical diabetics without liver disease upon some clinical features or the results of examination, and the results were as follows: 1. Of 119 cases of liver cirrhosis without primary diabetes 51 cases (42.9%) showed diabetic pattern in GTT, 36 cases (30.2%) probable diabetic, 7 cases (5.9%) oxyhyperglycemic, and 25 cases (21.0%) non-diabetic. 2. No incidence was found in the time of development or age distribution between the cases of liver cirrhosis with diabetic glucose tolerance and that combinated with primary diabetes or diabetic without liver disease. 3. Incidence of the abnormal glucose tolerance in the cases with liver cirrhosis did not correlate with the duration of the disease. 4. The abnormality of GTT showed high incidence in accordance with severity of liver injury in both of liver cirrhosis and hepatitis, however, it was impossible to find any significant correlation between the grade of abnormal glucose tolerance and the data of liver function test or the histological changes of the liver. 5. The glucose tolerance was not always aggravated inspite of progression from hepatitis to liver cirrhosis. The shift of the GTT and that of liver function disturbance were not always correlate with each other in both liver cirrhosis with abnormal glucose tolerance and that combinated with primary diabetes. 6. The grade or incidence of variou abnormalities from such points of view was high in accordance with the grade of abnormal glucose tolerance in cases with liver cirrhosis and no significant difference was seen as compared to the primary diabetics that the cases showing more than 100mg/dl of maximum fasting blood sugar in its duration, the cases with glycosuria before administration, or during GTT, occurance of the cases excreting more than 5 grams of glucose in whole day urine, influence of treatment with glucocorticoids, vacuolated nuclei occured in the liver cells, or development of complications related to the primary diabetes. It revealed that the stronger the abnormality of GTT in liver cirrhosis was, the more diabetic results. 7. The rate of abnormality of the results of prednisolonefrimed GTT was found to be higher in the cases of liver cirrhosis as compared to that of diabetes alone. 8. The incidence of occurence of vacuolated nuclei on the liver cells and of having complications related to primary diabetes was not different in the cases of liver cirrhosis with abnormal glucose tolerance or combinated with primary diabetes from that of chemical diabetes or clinical diabetes. 9. It might be possible to conclude from these results that the characteristics of abnormal glucose tolerance found in liver cirrhosis is similar to that found in primary diabetes and that the cases of liver cirrhosis with abnormal glucose tolerance might include primary diabetes or pancreas injury.