Aims: Few studies exist on influence of long-term oxygen treatment (LTOT) on survival of chronic obstructive pulmonary disease (COPD) patients. This study was designed to determine whether LTOT improves survival or not in severely hypoxaemic COPD patients. Materials: COPD patients prescribed oxygen concentrator were consecutively included. Patients' baseline characteristics were noted. During follow-up, patients were divided into three groups according to LTOT utilisation: (i) non-utilisers, (ii) intermittent utilisers (< 15 h/day) and (iii) true utilisers (15 h/day). Patients' status (live or death) and, if died, the date of death were checked throughout the study. The factors which might influence mortality during 5-year period were analysed. Results: Two-hundred and twenty-eight patients completed the study. Of these patients, 55 were in Group 1, 112 were in Group 2 and 61 were in Group 3. Regarding the initial characteristics, there was not any significant difference between groups. Mean follow-up for whole group was 27.8 +/- 18.5 months. Median survivals were similar between groups (19.5 +/- 5.6, 32.5 +/- 4.1 and 30.0 +/- 5.7 months respectively) (p > 0.05). Compared with Group 1, survival was improved in Group 2 (p < 0.05) and there was a positive trend for Group 3 during first 2-year period. However, this improvement disappeared during further follow-up. Analysis of multiple factors which might influence mortality during 5-year period did not yield statistically significant parameter. Discussion and conclusion: We found that, regarding survival, any kind of LTOT proved to be beneficial over no LTOT only in the first 2 years of follow-up, and that there was not any difference between intermittent and true LTOT utilisation.