BACKGROUND: Chemoembolization is often used in the treatment of hepatocellular carcinoma; however, there are limited data on its efficacy in an Australian setting. AIMS: To review retrospectively the experience of 21 patients with hepatocellular carcinoma who collectively had 36 chemoembolizations performed between October 1995 and February 1999 in a teaching hospital and liver transplant centre in Victoria. METHODS: Selective catheterization of the right or left hepatic arteries was performed. A mixture of cisplatin 50 mg, epirubicin 50 mg, mitomycin C 10 mg, Lipiodol and gelfoam was injected. Computed tomography (CT) scans were performed at baseline and at 1-3 months after chemoembolization. Outcome measures included response rates, toxicity, progression-free and overall survival. RESULTS: CT response rates: partial response 19% (n = 7), median duration 11 months (range 2+ to 37+); minor response 17% (n = 6), median duration 7 months (1+ to 12+); stable disease 42% (n = 15), median duration 3 months (1+ to 15 months); and progressive disease 22% (n = 8). Major toxicities included one case each of acute renal failure, contrast encephalopathy, gastric ulceration and hepatorenal failure. Median progression-free survival was 3 months (range 0-37+). Median overall survival was 15 months (range 6-50+). CONCLUSION: Chemoembolization has a role in the palliative treatment of hepatocellular carcinoma. Our response rates and toxicity data are consistent with those in the published literature. However, new treatments are needed and prevention of disease by reduction in the prevalence of chronic hepatitis B and C will be required to significantly reduce mortality from this tumour.