Accurate detection of hepatocellular carcinoma (HCC) foci in the liver and at the whole-body level has a significant impact on patient management. Functional whole-body imaging by PET (fused with CT or MRI) with spatial resolution compatible with the detection of lesions \2 cm in size has been proposed to overcome some limitations of morphological imaging. 18 F-fluorodeoxyglucose (FDG), the reference PET tracer in oncology, has limitations in the functional imaging of liver tumours. In particular , the detection rate of intra-hepatic, well-differentiated HCC is low and incompatible with effective staging of affected patients. To overcome this lack of sensitivity, choline PET tracers have been used by several teams: 11 C-choline or its analogues. 18 F-fluorocholine (FCH) and 18 F-fluoroethylcholine (FEC). These tracers showed sensitivity compatible with accurate staging of well-differentiated HCC and also of intermediate or poorly differentiated HCC. Dual-tracer PET using FDG and a lipid tracer has the best performance, since the aggressiveness of lesions within a given patient may vary, with some taking up only one tracer. Such variability of uptake may also be seen between different portions of a single large liver nodule. There is some evidence to suggest that a dual-tracer approach can be beneficial in the detection of distant metastases. Dual-tracer PET can also be useful in the selection of patients for liver transplantation or HCC tumour resection, for optimal pre-therapeutic staging, and potentially for prediction of recurrence. In pilot studies, visualisation of HCC tumours with FDG was found to indicate a worse prognosis, whereas visualisation with a lipid tracer was indicative of a better prognosis. Among non-HCC liver malignancies in adults, only cholangiocar-cinoma has been reported to take up lipid tracers in small series; FCH uptake has been reported in a child with recurrent hepatoblastoma. With regard to benign liver tumours, adenoma is rarely visible on choline PET, whereas focal nodular hyperplasia (FNH) is visible as a hot focus in the vast majority of cases. When using PET to characterise a liver nodule as HCC, this uptake by FNH may constitute a source of false-positive results. According to one team, FCH could be a good tracer to use in difficult cases for differentiating between FNH and hepatocellular adenoma which can potentially show malignant degenera-tion. From a logistical point of view, FCH is the easiest of the choline PET tracers: it has a longer half-life (110 min), can be produced industrially, and has been granted a marketing authorisation for this indication. The aim of the article is to provide an overview of PET imaging using the lipid tracer choline and its fluorinated analogues for studying HCC, summarising the currently available results. Color figures online at