Objective To evaluate the efficacy and safety of antiviral drugs for hepatitis B with YMDD motif variant. Methods We electronically searched MEDLINE (1989-April, 2004), EMBASE (1989-April, 2004), CBMdisc (expand) (1989-April, 2004), and handsearched unpublished Chinese conference proceedings. Randomized and quasi-randomized trials in patients with chronic hepatitis B with YMDD motif variant correlative to lamivudine were collected. Two reviewers extracted the data and assessed the quality of literature independently. The data were then analyzed by RevMan 4.2 software. Results Five studies involving 6 trials and 284 patients were included. According to the results of meta-analysis, antiviral therapy with adefovir plus lamivudine showed significantly better effects on the clearance of serum HBV-DNA and HBeAg and normalization of ALT than that of lamivudine alone (RR 16.61, 95%CI 2.29 to 120.71; RR 6.66, 95%CI 1.23 to 35.88 and RR 6.26, 95%CI 2.29 to 17.12 respectively); also, oxymatrine plus thymothin showed obviously better effects on the clearance of serum HBV-DNA and HBeAg (RR 2.96, 95%CI 1.26 to 6.93 and RR 2.51, 95%CI 1.05 to 5.98 respectively).But adefovir alone showed no better effects on clearance of serum HBV-DNA and HBeAg than that of lamivudine alone (RR 11.00, 95%CI 0.65 to 186.02 and RR 7.00, 95%CI 0.39 to 126.92 respectively); interferon plus lamivudine showed no better effects on the clearance of serum HBV-DNA, HBeAg and the normalization of ALT (RR 3.50, 95%CI 0.90 to 13.58; RR 4.90, 95%CI 0.70 to 35.10 and RR 2.80, 95%CI 0.91 to 8.12 respectively). Chinese herbs plus lamivudine showed no better effects on the clearance of serum HBV-DNA (RR 1.16, 95%CI 0.89 to 1.51). There were no significant side effects in the groups, except flu like symptom in the interferon group, slight kidney impairment in the adefovir group, and aggravation of rare cases in lamivudine group. Conclusions Antiviral therapy with adefovir plus lamivudine, or oxymatrine plus thymothin, shows better effects than with lamivudine alone in terms of antiviral therapy and clinical outcome improvement. However, the evidence is too weak to draw a definite conclusion in this systematic review. Larger sample size and rigorously designed randomized, double blind, placebo control trials are required for future study.