Background: In 1998, when data of a meta-analysis on tamoxifen in the treatment of hepatocellular carcinoma (HCC) had suggested a little advantage for this treatment, we published the results of a multicenter randomised controlled trial, that showed no survival benefit for tamoxifen vs. control. Here we report an updated analysis of the study results 4.5 years after the closure of enrollment. Methods: The study had a planned sample size of 480 patients. Patients with any stage HCC were eligible, irrespective of locoregional treatment. Tamoxifen was given orally, 40 mg/die, from randomisation until death. Results: 496 patients were randomised by 30 Institutions from January 1995 to January 1997. Information was available for 477 patients. As of July 2001, 374 deaths (78%) were recorded, and median survival times were 16 and 15 months (p=0.54), in the control and tamoxifen arm. Data were further analysed separately for advanced patients and for those eligible to potentially curative locoregional treatments: relative hazard of death for patients receiving tamoxifen was equal to 0.98 (95% CI 0.76-1.25) for the former group and 1.38 (95% CI 0.95- 2.01) for the latter. The prognostic score recently devised by our group (CLIP score) was, as expected, strictly correlated (p < 0.0001) to the locoregional treatment received and strongly correlated with prognosis. Conclusions: the update of the present study confirms that tamoxifen is not effective in prolonging survivals, both in advanced patients and in those potentially curable and that the CLIP score is able to predict prognosis.