Hepatocellular carcinoma (HCC) is one of the most severe sequelae of chronic liver disease. The only potentially curative therapeutic options are surgical resection and orthotopic liver transplantation. In most HCC patients, however, at clinical presentation the tumors are unresectable because of multicentricity or poor hepatic functional reserve due to pre-existing cirrhosis or not transplantable because of too advanced tumor stage or severe co-morbidity. In clinical practice, therefore, percutaneous ethanol injection (PEI) and transarterial chemoembolization (TACE) are widely used non-surgical therapeutic strategies. We prospectively analyzed the clinical factors determining the prognosis of 132 inoperable HCC patients and assessed the feasibility, therapeutic efficacy and safety of PEI, TACE and a combination thereof. Mean age of patients was 64 years; 95% of patients had liver cirrhosis and 39% were Okuda stage I, 48% stage II and 13% stage III. Fifteen patients were treated by PEI (group 1), 33 by TACE (group 2), 39 by TACE and PEI (group 3) and 45 received best supportive care (group 4). Survival correlated with the Child-Pugh class of liver cirrhosis and the Okuda stage of HCC. Favorable prognostic parameters were alpha-fetoprotein (AFP) levels <100 ng/ml and absence of portal vein thrombosis. Median survival time was 18 months in group 1 [interquartile range (IQR) 10-19], 8 months in group 2 (IQR 5-15), 25 months in group 3 (IQR 13-36) and 2 months in group 4 (IQR 1-9). Multivariate analysis revealed that patients treated with a combination of TACE and PEI have a significantly better survival than patients receiving either PEI or TACE only (p = 0.001). Patients with inoperable HCCs treated by the combination of TACE and PEI have a clear survival benefit. A favorable outcome can be expected in patients with compensated cirrhosis, a low Okuda stage, a baseline AFP level <100 ng/ml and absence of portal vein thrombosis.