Background: Postpneumonectomy empyema(PPE) is an infrequent but potentially life-threatening complication. To date, various surgical efforts have been made to manage this complication. We reviewed our 20-year surgical experience of PPE and long-term follow-up data. Material and Method: Total of 37 patients who were treated for PPE between fan, 1980 and Jun, 2000 were included. Various clinical factors such as micro-organism, operative method and timing, presence of bronchopleural fistula(BPF), underlying disease and fate of empyema cavity were retrospectively reviewed and analyzed. Result: Majority of patients(34) underwent Eloesser operation for effective drainage. There was only one operative mortality. The causative organisms were Staphylococcus species and Pseudomonas species in 46% BPF was found in 20 cases, among which spontaneous closures took place in 4 cases. The chest wall was closed in 40%(8/20) of patients with BPF, compared to 59%(10/17) without BPF. The closure rate was statistically better in patients without BPF(p=0.006). Even though the patients with benign disease showed higher closure rate(50%) than those with lung cancer (31%), the difference was not significant(p=0.25). Conclusion: Eloesser procedure was an effective method for initial drainage of PPE cavity with low operative mortality. Given the findings of low spontaneous closure rate of BPF, aggressive approach to close the BPF is mandatory to achieve the final goal of chest wall closure. It was found that majority of patients still left their chest cavity opened, even after controlling the active inflammation of the empyema cavity. More aggnessive approach for chest wall closure is recommended in all patents with benign disease and in selective patients with lung cancer if there is no evidence of recurrence at several years after the initial operation.