The effects of preoperative transcatheter arterial embolization (TAE) and intra-arterial injection of Lipiodol were retrospectively evaluated in patients with resectable hepatocellular carcinoma experienced during two different periods of time. In the TAE study, there were 31 patients with TAE and 107 patients without TAE. In the Lipiodol study, 60 patients had received Lipiodol injection and 68 patients served as controls. Curative hepatic resection was performed in all cases. As such preceding treatments had been performed at other hospitals, two comparing groups were fairly randomized in both the TAE and the Lipiodol study. Preoperative TAE did not influence the postoperative morbidity and mortality rates, cause of late death, and long-term survival rate in overall patients, but seemed to produce a better survival in noncirrhotic patients. Antitumor effect by Lipiodol injection was found in that the cancer free survival rate was significantly better in the group with Lipiodol injection. However, the overall 5-year survival rate was significantly better in the group without Lipiodol (67%) than in the group with Lipiodol (26%). The present study may indicate that preoperative TAE or Lipiodol injection should not be routinely performed. Such treatments may produce a substantial benefit only in selected patients with good hepatic functional reserve.