Objective To evaluate the effectiveness and safety of foscarnet sodium in the treatment of chronic hepatitis B. Methods We searched MEDLINE, EMbase, The Cochrane Library and CNKI from 1978 to June 2006. Randomized controlled trials of foscarnet sodium versus other drugs or no drugs in the treatment of chronic hepatitis B were identified. The quality of the included trials was evaluated by two reviewers independently. Meta-analysis was done using The Cochrane Collaboration’s RevMan 4.2.7. Results Seven studies (337 patients) were included; one compared foscarnet sodium versus interferon, and the other six compared foscarnet sodium versus no drugs. All the included studies were graded in terms of the quality of randomization, allocation concealment and blinding. All 7 studies were graded as level C. The meta-analysis showed that: ① foscarnet sodium was not significantly different from interferon in clinical efficacy, liver function, negative-conversion rate of virological markers and side effects. ② compared with the no drugs group, the negative-conversion rate of virological markers was significantly higher for the foscarnet sodium group, HBeAg (RR 6.20, 95%CI 1.76 to 21.79) and HBV-DNA (RR 4.13, 95%CI 1.32 to 12.86); but there were no significant differences in clinical efficacy, liver function and side effects. Conclusions Available evidence shows that: in the treatment of chronic hepatitis B the effectiveness and safety of foscarnet sodium are not significantly different from interferon, but only one trial is included in this review, so the evidence is weak. Compared with no drugs, foscarnet sodium significantly improves the negative-conversion rate of virological markers, but the evidence is insufficient to show whether foscarnet sodium could improve clinical efficacy and liver function, as well as reduce side effects.