Hepatic assist remains elusive. Bioartificial livers (BALs), consisting of liver cells or tissue in a synthetic housing, have been promising but have not proven successful in clinical trials. Artificial livers that consist of sophisticated sorbents and membranes cannot support a failing liver but may shorten episodes of acute decompensation in patients with stable cirrhosis. These artificial livers are most likely to find a place as temporary support prior to transplantation. True liver support will require a BAL. This article proposes goals for making a clinically useful BAL, with attention to systems biology and potential sources of hepatocytes.