In 481 patients, with sonographically suspected neoplastic involvement of the liver, we performed an ultrasonically guided fine-needle aspiration biopsy with 22-gauge Chiba needle. In the 441 controlled cases, we obtained 68.5% true positive, 26.5% true negative, and 5% false negative, with a sensitivity of 93.2%, specificity of 100%, and overall accuracy of 95%. The true positive include 42.1% hepatocellular carcinomas, 28.1% metastases from known extrahepatic malignancy, 24.5% metastases from unknown primary site, 5% hepatic involvement by non-Hodgkin lymphomas, and 0.3% undetermined malignancy. Also, in a larger series, ultrasonically guided fine-needle biopsy confirms its high specificity and overall accuracy. The cytologic assessment of the obtained material is almost always adequate. The use of immediate cytologic staining reduces the number of passes in each lesion with absence of insufficient sampling. Biopsy with fine cutting needles should be reserved only for cases not diagnosed by cytology.