Smoot and co-workers report on a large retrospective analysis of patients with hepatocellular carcinoma (HCC) without cirrhosis from a single institution. HCCs in non-cirrhotic liver are not rare; if expressed as a percentage of all HCCs in a geographical area, their proportion ranges between 7 and 54 per cent1. Most institutes only occasionally encounter HCC in a background normal liver, and so published data are relatively scarce. The study of Smoot and colleagues adds additional data to our knowledge about HCC in non-cirrhotic liver and might contribute to future treatment strategies. In clinical practice the first challenge is to diagnose HCC without cirrhosis correctly and the second step is to decide on resectability. Using computed tomography and modern liver function tests this can presently be done in one investigation, if necessary2. In this context, a major issue concerns the long-term oncological outcome and survival. Should all eligible patients be resected or is there a group that should undergo primary liver transplantation? In this article Smoot et al. highlighted that the presence of two or more tumours resulted in an increased risk of recurrence and reduced survival. Moreover, 75 per cent of recurrences occurred exclusively in the liver. Thus, one could hypothesize that patients with HCCs without cirrhosis and two or more tumours fulfilling (extended) accepted criteria should undergo transplantation primarily, regardless of resectability. The latter patient group might have a generally increased risk of developing HCC of unknown cause (genetics, toxins?) or on the basis of underlying chronic liver disease (up to 30 per cent in this article)1. Thus, this group might directly benefit from such a strategy. Alternatively, initial liver resection with later salvage transplantation could be performed to single out patients with worse tumour biology, if additional tumours are present resembling metastatic disease. On the other hand, salvage transplantation is associated with a poorer outcome than primary liver transplantation3. In the end, to clarify these issues only large multicentre multinational studies or a register could increase the evidence.