It is essential to estimate the liver size in the diagnosis of liver disease. Many approaches have been attempted in the evaluation of liver size such as measurement of length, area and volume. Among these, are and volume measurements are accurate but complicated. So various linear measurements including Pfahler's method have been commonly used. But to our knowledge, there was no report about linear diameter of liver in Korean adults. As in general, larger patients are likely to have larger liver than smaller ones. it seems to be more ideal to evaluate the liver size using relative ratio rather than absolute linear diameters. The main objectives of our investigation were to determine the various diameters of normal and enlarged liver and the criteria of hepatomegaly in Korean adults using absolute and relative ments. Our cases consisted of 95 clinically normal subjects and 51 patients suffering from liver idsease and diagnosed to have hepatomegaly on abdominal palpation and simple abdomen in the Dept. of Radiology, St. Mary Hospital during the period of 6 months since Jan. 1981. We measured the liver size using 3 liner diameters (Fig. 1). And as the refernce masurement, the distance from the right margin of the liver to the left margin of spleen was also measured. We called this "abdominal transverse diameter" (ATD). The results were as follows : 1. The diameters of liver were 13.4$\pm$1.6cm, 18.4$\pm$2.4cm, 19.2$\pm$2.6cm in normal group and 18.8$\pm$3.1cm, 23.5$\pm$3.0cm, 24.2$\pm$3.2cm in hepatomegaly group using midline vertical diamter (MEVD), maxim vertical diameter (MAVD) and diagonal diameter (DD), respectively. The difference between two groups were statistically very significant in every method (p < 0.01) 2. The 99% tolerance limits of liver diameters were 13.0$\pm$13.8cm, 17.8$\pm$19.0cm, 18.5$\pm$19.9cm in normal group and 17.7$\pm$19.9cm, 22.4$\pm$ 24.6cm, 23.1$\pm$25.3cm in hepatomegaly group using MIVD, MAVD and DD, respectively. The midpoints between the upper limit of normal group and the lower limit of of gepatomegaly group were 15.7cm, 20.7cm, 21.5cm in the case of MIVD, MAVD and DD, respectively. These points are warranted to suggest criterias of hepatomegaly. 3. There were statistical significant difference in the ratio of each diameter to ATD between normal and hepatomegaly group (p < 0.01). We called this "hepato-abdominal ratio". The "hepato-abdominal ratio" were 0.44$\pm$0.05, 0.61$\pm$0.07, 0.64$\pm$0.08 in normal group and 0.61 $\pm$0.10, 0.76$\pm$0.08, 0.78$\pm$0.09 in hepatomegaly group using MIVD, MAVD and DD, respectively. 4. The 99% tolerance limits of "hepato-abdominal ratio" were 0.43-0.45, 0.59-0.63, 0.62-0.64 in normal group and 0.60-0.62, 0.75-0.77, 0.77-0.79 in hepatomegaly group using MIVD, MAVD and DD, respectively. The midpoints between the upper limit of normal group and the lower limit of hepatomegaly group were 0.52, 0.69, 0.70 using MIVD, MAVD and DD, respectively. These points are also warranted to suggest criterias of hepatomegaly. to suggest criterias of hepatomegaly.