The Milan criteria have been proven to be reliable and easily applicable in selection of patients with small unresectable hepatocellular carcinomas for liver transplantation. It has been repeatedly shown that patients who met these criteria had a 5-year survival of over 70％ after transplantation. Such a result is remarkably good for an otherwise incurable malignancy. The main disadvantage of this set of criteria is that it is rather restrictive. Following it religiously denies transplantation to many patients who have tumor stage slightly more advanced. There have been many attempts to extend the criteria to include tumors with larger sizes (as in the UCSF criteria) or with a larger number (as in the Kyoto criteria). Alpha-fetoprotein and PIVKA-II, two biological markers in more aggressive tumors, have also been employed in the selection of patients, and biopsies have been used by the University of Toronto to determine tumor aggressiveness before deciding on transplantation. Patients with tumors beyond the Milan criteria yet not of a high grade have been accepted for transplantation and their survival is comparable to that of transplant recipients who were within the Milan criteria. Preoperative dual-tracer (11C-acetate and FDG) positron emission tomography has been used to determine tumor grade, and transarterial chemoembolization has been used to downstage tumors, rendering them meeting the Milan criteria. Patients with downstaged tumors have excellent survival after transplantation. Partial response to chemical treatment is a reflection of less aggressive tumor behavior. Careful selection of patients beyond the Milan criteria with the aid of serum tumor marker assay, positron emission tomography or tumor biopsy allows transplanting more patients without compromising survival. The use of liver grafts either from the deceased or from living donors could thus be justified.