Objective To analyze the treatment and effect of bacterial liver abscess over the past two decades in one single center. Methods The total 198 patients with bacterial liver abscess during the last twenty years were studied retrospectively. They were divided into three groups according time: 1989-1995 group, 1996-2002 group and 2003-2008 group. Gender and age of patient, location, number and size of abscesses, treatment, hospital days, morbidity of complications and mortality among the groups were compared. Results There were 54, 69, 75 cases in 1989-1995, 1996-2002 and 2003-2008 group respectively. No significant differences were found in gender and age of patient, location, number and size of abscess among three groups （Pgt;0.05）. In 1989-1995 group, 35 cases (64.8%) were treated with laparotomy, 8 cases (14.8%) with laparoscope, and 11 cases (20.4%) with percutaneous treatment (needle aspiration or catheter drainage). In 1996-2002 group, 15 cases (21.8%) were treated with laparotomy, 21 cases (30.4%) with laparoscope, 31 cases (44.9%) with percutaneous treatment (needle aspiration or catheter drainage), and 2 cases (2.9%) were treated with antibiotherapy. In 2003-2008 group, 5 cases (6.7%) were treated with laparotomy, 13 cases (17.3%) with laparoscope, 54 cases (72.0%) with percutaneous treatment (needle aspiration or catheter drainage), and 3 cases (4.0%) were treated with antibiotherapy. The constituent ratio of treatment was significantly different among three groups (P＜0.05). The hospital days was (18.5±12.2) d, (16.4±12.8) d and (20.1±14.6) d, the morbidity of complications was 9.3% (5/54), 4.3%(3/69) and 4.0%(3/75), the mortality was 3.7%(2/54), 1.4%(1/69) and 1.3% (1/75) respectively, but there were no significant differences of three indexes among three groups. Conclusion With the development of surgical techniques, effective antibiotic therapy and percutaneous treatment (needle aspiration or catheter drainage) have been the main therapeutic methods, and laparoscopy and laparotomy are necessary supplement.