Prognostic factors in COPD patients receiving long-term oxygen (LTO) therapy were recently analyzed, but very few studies considered the prognostic value of pulmonary artery pressure (PAP) in these patients. We investigated 84 patients who had undergone a right heart catheterization just before the onset of LTO. There were 75 men and 9 women, with a mean age of 63.0 +/- 9.9 (SD) years, at the onset of LTO. When PaO2 was persistently less than 55 mm Hg, LTO was initiated. This therapy was started in some patients with PaO2 in the range of 55 to 60 mm Hg if they had signs of cor pulmonale or a resting PAP of 25 mm Hg or greater at right heart catheterization. The daily duration of LTO was 16 h/d or more. Oxygen flow was adapted to achieve a PaO2 of 65 mm Hg or more. The patients were subdivided into subgroups according to the median value of age (cutoff value = 63 years); vital capacity (2,250 mL); FEV1 (800 mL); residual volume-total lung capacity ratio (58%); PaO2 value (52 mm Hg), PaCO2 level (45 mm Hg); and PAP (25 mm Hg). The cumulative 5-year survival rate was 48% for the group as a whole. Actuarial survival curves were plotted for the two subgroups of patients subdivided according to the initial median value of the variables just listed. There was no significant difference in survival rate between subgroups except when taking into account the level of PAP and age. In patients with an initial PAP of 25 mm Hg or less (n = 44), the 5-year survival was of 62.2 vs 36.3% in the remainder (n = 40) [p < 0.001]. We performed a multivariate analysis of survival using Cox's model of the proportional hazards regression including sex and the variables with the same categorization in the stepwise procedure: PAP and age were the only variables included in the final model. We conclude that the best prognostic factor in COPD patients receiving LTO is not the FEV1, nor the degree of hypoxemia or hypercapnia, but the level of PAP.