BACKGROUND:: It has been suggested that lamivudine therapy may be even more effective if administered at higher doses than is dictated by the current standard regimen. We analyzed the correlation between the effects of lamivudine and body surface area (BSA). METHOD:: We evaluated 134 patients with chronic hepatitis B who had been treated with lamivudine for more than 12 months. The effect of the treatment was evaluated from the levels of serum alanine aminotransferase (ALT) and HBV-DNA. Several variables that could influence the response to treatment, including ALT, albumin, and bilirubin levels, platelet counts, BSA, HBV-DNA, and HBeAg were analyzed. RESULTS:: Univariate logistic analysis selected platelet counts, BSA, HBV-DNA and HBeAg in the biological evaluation, and bilirubin, BSA, HBV-DNA and HBeAg in the virological evaluation (chi(2)>1.0). Using these factors, multivariate analysis revealed that BSA (chi(2)=12.8, p=0.0004) was the only factor that could contribute significantly to the improvement of ALT levels, and that BSA (chi(2)=4.4, p=0.0354) and HBeAg (chi(2)=8.1, p=0.0044) were independent factors that could influence the suppression of HBV-DNA. CONCLUSION:: We revealed that BSA is a significantly predictor of the effect of lamivudine therapy, suggesting that lamivudine dosage should be based on the individual BSA.