Long term oxygen therapy (LTOT) has been shown to improve the survival rate in Chronic Obstructive Pulmonary Disease (COPD) patients with severe resting hypoxemia by NOTT and MRC studies, published more than 25 years ago. The improved survival was found in patients who received oxygen for more than 15 hours/day. The effectiveness of LTOT has been documented only in stable COPD patients with severe chronic hypoxemia at rest (PaO2 < 55 mmHg (7.3 kPa) or PaO2 ranging from 56 to 59 mmHg (7.4-7.8 kPa) in presence of signs of Cor Pulmonale, hematocrit > 55%. In fact no evidence supports the use of LTOT in COPD patients with moderate hypoxemia (55 < PaO2 < 65 mmHg), and in those with decreased oxygen saturation (SO 2 <90%) during exercise or sleep. Furthermore, it is generally accepted without evidence that LTOT in clinical practice is warranted in other forms of chronic respiratory failure not due to COPD when a-terial blood gas criteria match those established for COPD patients. The prescription of oxygen in these circumstances, as for unstable patients, increases the number of patients receiving supplemental oxygen and the related costs. Comorbidities are likely to affect both prognosis and health outcomes in COPD patients, but at the moment we do not know if LTOT in these patients with complex chronic diseases and mild-moderate hypoxemia could be of any use. For these reasons a critical revision of the actual guide lines indications for LTOT in order to optimise effectiveness and costs, and future research in the areas that have not previously been addressed by NOTT and MRC studies, are mandatory.