Many treatments have been proposed for non-resectable primary or secondary hepatic cancer but the results have generally been disappointing. Isolated Hepatic Perfusion (IHP) was first attempted four decades ago but it gained acceptance only recently, after spectacular tumour responses were obtained by isolated limb perfusion with melphalan and tumour necrosis factor (TNF) for melanomas and sarcomas. Surgical isolation of the liver is a technically demanding operation that allows the safe administration of high doses of chemotherapeutics and TNF. Percutaneous techniques using balloon occlusion catheters are simpler but result in higher leakage rates from the perfusion circuit into the systemic circulation. Several phase I-II trials indicate that IHP can yield high tumour response rates, even when there is resistance to systemic chemotherapy. However, no significant advantage in overall survival has been demonstrated so far. IHP offers unique pharmacokinetic advantages for locoregional chemotherapy and biotherapy. It might also allow gene therapy with limited systemic exposure and toxicity. At present, IHP nevertheless remains an experimental treatment modality which should therefore be used in controlled trials only.