The hepatitis B virus (HBV) is the world's most important chronic virus infection. The immunomodulator interferon-alpha (IFN alpha) is the only clinically applied drug available, despite its low response rate (approximately 30%) even in highly selected chronic carriers. Antiviral nucleoside analogues have proven to be potent inhibitors of viral replication in vitro, but their significant adverse effects which are, at least partially, due to their nonspecific body distribution, have forced the cessation of their clinical development in the past. For example, vidarabine causes severe neuromuscular toxicity, and fialuridine has caused fatal cases of liver and kidney failure in a recent clinical trial. Furthermore, the potential clinical application of (modified) antisense oligodeoxynucleotides, which are very specific inhibitors of viral replication, is hampered by their nonspecific body distribution, instability in serum and poor cell penetration. As infection and replication of HBV mainly occur in liver parenchymal cells, selective targeting of antiviral nucleoside analogues as well as antisense oligodeoxynucleotides to the liver would theoretically improve therapeutic efficacy. At present, conjugates of vidarabine and neoglycoproteins have entered clinical trials, and initial data suggest that therapeutic concentrations are achieved at lower dosages with minor adverse effects. These data have stimulated preclinical research on other liver-specific drug carriers for the selective delivery of HBV-active drugs such as glycosylated polymers and neolipoproteins: these approaches are outlined in this paper.