Chronic obstructive pulmonary disease (COPD) is a common disorder that affects approximately 400,000 Danish citizens. About 3,000-3,500 Danes die yearly because of the disorder, and the costs associated with hospital admissions are estimated to be 535 million Danish kroner (DKK). Patients with COPD risk a worsening of their disorder, and in most cases, this will require hospitalization. One of the used treatments is providing oxygen to the patients via e.g. masks. The recommendations on oxygen treatment are currently based on a study from 2010 where 37% of the participants in this study did not receive the intended treatment, which may have had massive effects on the results. It is worrying that no other studies have shown which oxygen treatment is safest for the patients. As such, we deem it important to study how best to treat the patients. Our study is of high clinical relevance as hospitals receive patients with worsening of COPD daily. We need more, better data regarding the oxygen treatment of our patients, in order to provide our patients with the best possible care. The purpose of our study is thus to determine which oxygen treatment is best for patients with acute worsening of COPD symptoms. We will use a prospective, randomized controlled open-label trial. We will use two treatments: Treatment 1 is giving oxygen to the patient to reach a peripheral oxygen saturation of above 94%. Treatment 2 is giving oxygen to reach a peripheral oxygen saturation of between 88% and 92%. Our primary outcome is 30-day all-cause mortality, with secondary outcomes being 7-day all-cause mortality, need for non-invasive ventilation, intubation or intensive care admission, over-all length of hospital stay and respiratory acidosis. We believe that a lower oxygen saturation percentage may be superior as one study (Austin et al., 2010) showed a lower mortality rate in the group of patients that had a lower peripheral oxygen saturation. Additionally, the risk of respiratory acidosis and hypercapnia were lower. We wish to perform our study in the hospital sector as this study was performed in the prehospital sector and thus their results cannot be translated directly.