The prognostic value and the evolution of pulmonary hypertension (PH) in patients with markedly hypoxemic chronic obstructive pulmonary disease (COPD), treated or not with long-term oxygen therapy (LTOT), has been extensively investigated. However, little is known in patients with mildly or moderately hypoxemic COPD not requiring LTOT. Therefore, we assessed the evolution of pulmonary hemodynamics in 131 patients with stable COPD by performing two right heart catheterizations at a mean (+/- SD) time interval of 6.8 +/- 2.9 yr. At inclusion (T0), no patient had PH (i.e., the mean pulmonary artery pressure [Ppa] at rest was < 20 mm Hg). Group 1 included 55 patients without exercising PH and group 2 included 76 patients with exercising PH, defined by a pulmonary arterial pressure (Ppa) > 30 mm Hg during a steady-state 40-W exercise. Group 2 patients compared with group 1 patients had a significantly higher resting Ppa (16 +/- 3 mm Hg versus 14 +/- 2 mm Hg, p = 0.001). At the second catheterization, 33 (25%) patients (9 of 55 in group 1, 24 of 76 in group 2, p = 0.048) showed a resting Ppa > 20 mm Hg, but PH was generally mild, ranging from 20 to 42.5 mm Hg. The mean Ppa at second evaluation was 16 +/- 5 mm Hg in group 1 and 19 +/- 7 mm Hg in group 2 (p = 0.01). The patients who developed resting PH at the second catheterization (T1) had higher resting and exercising Ppa (p = 0.001 and p = 0.002, respectively), and significantly lower resting and exercising Pa(O(2)) (p = 0.005 and p = 0.012, respectively) at T0. Logistic regression analysis showed that resting and exercising Ppa were independent predictors (at T0) for the subsequent development of PH (p = 0.029 and p = 0.027, respectively). The patients who developed resting PH (T1) had a significantly worsening of Pa(O(2)) (from 63.5 mm Hg at T0 to 60 mm Hg at T1, p = 0.047), whereas the Pa(O(2)) as a mean was stable in the remainder (69.5 mm Hg at T0 and T1). These results show the following. The progression of Ppa over time in patients with COPD with mild to moderate hypoxemia is rather slow, the average change for the group as a whole being of + 0.4 mm Hg/yr. Only about 25% of patients with COPD with mild to moderate hypoxemia and without resting PH at the onset will develop PH during a 6-yr follow-up. The patients with exercising PH at the onset have a significantly increased risk of developing PH over time. Only resting and exercising Ppa at the onset are independently related to the subsequent development of PH. However, in individual cases, the models of linear or logistic regression do not allow a pertinent prediction of the level of Ppa or the presence of PH at the second right heart catheterization.