Severe coagulation defects, as reflected by platelet count and prothrombin time, have always been considered a contraindication to needle biopsy of the liver, but there are very limited data on the actual rate of bleeding in patients with such severe alterations and none whatsoever on the bleeding risk associated with newer, fine-gauge needles that produce less trauma to the liver tissue. In addition, there has never been any evidence that platelet count and/or prothrombin time are the most sensitive indices of bleeding risk. This retrospective study of 85 patients, with platelet counts less than 50,000/mm3 and/or prothrombin times less than 50% of controls, subjected to ultrasound-guided fine-needle liver punctures for diagnostic or therapeutic (percutaneous ethanol injection) purposes showed no bleeding episodes after any of the 229 punctures performed. No type of replacement therapy was administered to correct clotting defects prior to the procedure. Correct pathologic diagnoses were obtained in 81.2% of all patients. Ultrasound-guided fine needle puncture appears to be safer than currently believed in patients with severe clotting defects and deserves further evaluation as an alternative to surgical procedures to diagnose and treat liver lesions, even when severe coagulation impairment is present.