Background: The aim of the study was to apply a computer aided analysis of the three-dimensional (3D) ultrasound images for the diagnosis of focal liver lesions. Material/Method: Patients with various hepatic changes were examined using standard US unit equipped with a transducer positioning system based on the magnetic field sensor. The data were imported by the graphical workstation and processed. The computer analysis comprised 3D image segmentation as well as the lesion volume and echogeneity determination. Moreover, a set of parameters was calculated to quantify morphology and texture of the lesions. Results: The collected results confirmed that calculations of lesion volume and echogeneity as well as quantification of the morphological features have a limited value in the diagnosis of liver tumors. In contrast, the quantification of selected texture features of the 3D liver images enables differentiation between malignant and benign tumors. The best parameters to distinguish between different types of the liver lesions are angular second moment and entropy. Conclusions: The results suggest that quantification of the texture features offers the supplementation of the routine US approach to the liver lesion diagnosis by a new quantitative tool.