We systematically reviewed the evidence for determining the best radiological imaging for characterizing hepatocellular carcinoma (HCC) in cirrhotic patients in 997 articles between 1995 and 2001. We selected only prospective and retrospective cohorts of patients, excluding both case reports and studies without separate data on HCC. Only 29 studies, comprising 918 patients, fulfilled the inclusion criteria: 10 used the explanted liver as the reference standard of diagnosis. All except one, either found no statistically significant difference between imaging modalities or had no direct comparison of sensitivity between different modalities of imaging; 16 studies evaluated HCC among cirrhotic patients and had biopsy or imaging as the reference standard for diagnosis. However, no one imaging technique was shown to be superior. In two studies, data of a HCC subgroup was derived from the studies evaluating different kinds of focal hepatic lesions. No conclusion could be drawn because of the small sample size. One study addressed the issue of therapeutic impact. The evidence for choosing the best modality of imaging for characterizing HCC in cirrhotic patients is inadequate. Large multicentre studies with defined reference standards for diagnosis, and studies evaluating therapeutic impact are needed.