AIM: To further evaluate the relationship between BSA and the effects of lamivudine in a greater number of cases and over a longer period of observation than in our previous evaluation.METHODS: We evaluated 249 patients with chronic hepatitis B. The effects of treatment for one year (n = 249),two years (n = 147), and three years (n = 72) were evaluated from the levels of serum ALT and HBV-DNA,as biological and virological effects (undetectable levels by PCR), respectively. Moreover, several variables that could influence the response to treatment, including ALT,albumin, bilirubin, platelet counts, BSA, HBV-DNA, and HBeAg were analyzed.RESULTS: For 1-year treatment, multivariate analysis revealed that BSA (P=0.0002) was the only factor for the biological effect, and that ALT (P = 0.0017), HBVDNA (P = 0.0004), and HBeAg (P = 0.0021) were independent factors for the virological effect. For 2-year treatment, multivariate analysis again showed that BSA(P = 0.0147) was the only factor for the biological effect,and that ALT (P = 0.0192) and HBeAg (P = 0.0428) were independent factors for the virological effect. For 3-year treatment, multivariate analysis, however, could not reveal BSA (P = 0.0730) as a factor for the normalization of ALT levels.CONCLUSION: BSA is a significant predictor for the normalizing the effect of lamivudine therapy on ALT for an initial 2-year period, suggesting that lamivudine dosage should be based on the individual BSA.