This study was intended to compare the survival rates of two contemporary cohorts of patients with solitary hepatocellular carcinomas < or = 4 cm subjected to surgical resection (n = 33) or percutaneous ethanol injection (n = 30). Outcomes in a third cohort, 21 patients with hepatocellular carcinoma who underwent orthotopic liver transplantation, were also assessed. Surgical and ethanol-treated patients were similar with regard to age and tumor stage, differing only in liver function; 30 of the 33 surgical patients were of Child-Pugh class A, whereas only 7 of the 30 ethanol-treated patients were of class A (p < 0.05). Surgical resection was successful in 30 cases; ethanol injection achieved initial success in 23 patients. Tumor recurrence rate at 2 yr was 45% in the surgical group and 66% in the ethanol group. The difference was significant only for cases with tumors between 3 and 4 cm. Despite poorer liver function, the 1-, 2-, 3- and 4-yr survival rates of ethanol-treated patients (83%, 66%, 55% and 34%, respectively) were not different from those of surgically treated patients (81%, 73%, 44% and 44%, respectively). The 1- and 2-yr survival rates of patients given liver transplants were 81% and 66%, without tumor recurrence, after 16-mo follow-up. These data confirm that ethanol injection is a useful treatment for patients with solitary small hepatocellular carcinomas and suggest that surgical resection and liver transplantation may achieve better results only after strict candidate selection to reduce mortality and tumor recurrence during follow-up.