AIM: To study the anti-HBV effect of liposome-encapsulated matrine (Lip-M) in vitro and in vivo.METHODS: 2.2.15 cell line was cultured in vitro to observe the effect of Lip-M and matrine on the secretion of HBsAg and HBeAg. The toxicity of Lip-M and matrine to 2.2.15 cell line was also studied by MTT method. In in vivo study,drug treatment experiment was carried out on the 13th day after ducks were infected with duck hepatitis B virus (DHBV). The ducks were randomly divided into 4 groups with 5-6 ducks in each group. Lip-M and matrine were given to DHBV-infected ducks respectively by gastric perfusion. Four groups were observed: group of Lip-M (20 mg/kg), group of Lip-M (10 mg/kg), group of matrine (20 mg/kg) and group of blank model. The drug was given once daily for 20 d continuously, and normal saline was used as control. The blood was drawn from the posterior tibial vein of all ducks before treatment (T0), after the medication for 5 (T5), 10 (T10), 15 (T15), 20 (T20) d and withdrawl of the drug for 3 d (P3). The serum samples were separated and stored at -70 ℃, DHBV-DNA was detected by the dot-blot hybridization.RESULTS: After addition of Lip-M and matrine to 2.2.15cell line for eleven d, the median toxic concentration (TC50)of Lip-M and matrine was 7.29 mg/mL and 1.33 mg/mL respectively. The median concentration (IC50) of Lip-M to inhibit HBsAg and HBeAg expression was 0.078 mg/mL and 3.35 mg/mL respectively. The treatment index (TI)value of Lip-M for HBsAg and HBeAg was 93.46 and 2.17respectively, better than that of matrine. The DHBVinfected duck model treatment test showed that the duck serum DHBV-DNA levels were markedly reduced in the group of Lip-M (20 mg/kg) after treated by gastric perfusion for 10, 15 and 20 d (0.43±0.22 vs0.95±0.18,t= 4.70, P= 0.001＜0.01.0.40±0.12 vs0.95±0.18, t= 6.34,P= 0.000＜0.01. 0.22±0.10 vs 0.95±0.18, t= 8.30,P = 0.000＜0.01), compared to the group of matrine (20 mg/kg) (0.43±0.22 vs0.79±0.19, t= 3.17, P= 0.01＜0.05.0.40±0.12 vs0.73±0.24, t= 3.21, P= 0.009＜0.05.0.22±0.10vs0.55±0.32, t= 2.27, P = 0.046＜0.05.), and the control (0.43±0.22 vs 0.98±0.29, t = 3.68, P = 0.005＜0.01.0.40±0.12 vs0.97±0.30, t= 4.26, P= 0.002＜0.01.0.22±0.10vs0.95±0.27, t= 5.76, P= 0.000＜0.01). After the treatment for 20 d and withdrawl of the drug for 3 d, duck serum DHBV-DNA level in the group of Lip-M (10 mg/kg) markedly reduced (0.56±0.26 vs 0.95±0.38, t = 5.26, P = 0.003＜0.05.0.55±0.25 vs 0.95±0.38, t = 5.52, P = 0.003＜0.05), and the difference was significant as compared with the control (0.56±0.26 vs 0.95±0.27, t = 2.37, P = 0.042＜0.05.0.55±0.25 vs0.89±0.18, t= 2.55, P= 0.031＜0.05), but not significant as compared with the group of matrine (20 mg/kg).After withdrawl of the drug for 3 d, the levels of DHBV-DNA did not relapse in both groups of Lip-M.CONCLUSION: Lip-M can evidently inhibit the replication of hepatitis B virus in vitro and in vivo; its anti-HBV effect is better than that of matrine.