Grayscale ultrasonic imageing allows a detailed analysis of the contour and echotexture of internal organs and thus the characterisation of both diffuse pathological processes and space- occupying lesions. In order to evaluate the reliability of the uItrasonography in the diagnosis of diffuse parenchymal liver diseases, abdominal ultrasonograms of 45 subjects who undeirwent liver biopsy within 2 Weeks after ultrasonic examination were investigated retrospectively. On the histopathologic findings, 5 out of 45 cases were normal. The remaininy 48, however, had parenchymal disease proceses:acute viral hepatitis 8 (type A 3, type B 5), chronic hepatitis 9 (chronic persistent hepatitis 8, chronic active hepatitis 1), fatty liver 9. And liver cirrhosis with or without other pathologic process l4, respectively. In this study, a scoring system is designed to estimate the possible role as objective diagnostic criterlor: on the basis of combination of each parameter in the patients with parenchymal liver diseases. Each patient was scored on the basis of summing up the points weighting as 1 point for each finding of 11 parametes hepatomegaly, inhomogeneous texture, accentuation of intenal echogenicity, nodular echo, intrahepatic vascular distortion, diminished beam penetration. To the distal deep area, downward convexity of lower edge, marginal hump, irregularity and the anterior hepatic surface, spienomegaly, and ascites. Mean score was marked as 0.4+-0.9 (in the normal control cases, 2.4+-1.4 in group I (acute viral hepatitis), 3.4+-1.5 in II (chronic hepatitis), 5.2+-1.9 in III (fatty liver), and 6.0+-2.1 in IV (liver cirrhosis), respectively, In view of ultrasonograms inability to distinguish normal livers from thcse with hepatitis or fatty from cirrhotic livers significantly, it was convinient to combine the normal group with groups I and II, and groups III and IV, for further analysis. It was readily apparen t that 73.9% in sensitivity, 81.8% in specificity, 18.2% in false positivity and 26.1% in false negativity with 80.9% in diagnosability for detecting fatty and or cirrhotic liver were obtained when the scores were remarked between 5 to 10 Further weightirg to ancillary findings such as splenoimegaly and ascite., and hepatic images of nodular Echo. Hump and marginal irregularity were meaningful in reaching a correct differential diagnosis between fatty and cirrhotic liver.