The development of HCV subgenomic replicons that produce high levels of one or more HCV polypeptides (Lohmann et al., 1999; Blight et al., 2000), in contrast to the very low and inconsistent levels of wild-type virus produced in tissue culture cells infected with HCV (Table 11.1 and Ch. 11), suggests that there are properties of wild-type HCV that normally attenuate virus gene expression and replication, and that when these constraints are removed, much higher levels of expression could be achieved. One of these constraints may be the complex secondary structural features within the 5' and 3' UTRs of the virus. Many laboratories are developing additional self-replicating replicons that are capable of persisting, and some at high copy number, within transfected or infected cells. These include replicons made from alphavirus (Garoff & Li, 1998; Ying et al., 1999), pestivirus (Moser et al., 1999), other flavivirus (Varnavski & Khromykh, 1999), and coronavirus (Thiel, Siddell & Herold, 1998) vectors. Whether the entire HCV polyprotein could be expressed from such constructs remains to be seen. In addition, since the HCV would be produced from artificial templates, it is not clear whether the sensitivity of virus gene expression and replication to putative antiviral agents would be the same or different to that of virus made from native HCV templates in an infected cell.