Patients with pulmonary fibrosis experience early oxyhemoglobin desaturation under effort, which limits their ability to exercise and their quality of life. Recent studies have shown that in resting normoxaemic patients who become hypoxemic under exertion, administration of outpatient oxygen significantly improves stress dyspnoea and quality of life. It is unclear how this happens, since oxygen administration does not act directly on dyspnoea, and does not appear to have much effect on the heart rate and pulmonary artery pressure. We tested the hypothesis that correcting the hypoxaemia could reduce the increase in respiratory effort during the 6 min walking test, recording the breathing pattern during administration of oxygen or placebo. We evaluated 20 patients with fibrotic interstitial lung diseases (17 males and 3 females; mean age 72 ± 2 years; M ± SE) with a resting SpO2 ≥92 that fell to ≤88% during the 6 min walk test (6MWT). After first establishing the oxygen flow necessary to prevent desaturation, the patients underwent two further 6MWT, 15-20 min apart, one with administration of medical air and one with oxygen at the same flow, in randomized double-blind order. During the test, SpO2, heart rate, respiratory rate, tidal volume and minute ventilation (VE) were recorded, using a Spiropalm spirometer (Cosmed, Rome, Italy). Oxygen saturation during the 6MWT decreased to a minimum value of 82.3% (95% CI 80.1-84.5%) during placebo and to 92% (90.3-93.7%) during oxygen with an average difference of 9.7% (7.8-11.6%, p < 0.0001). On the contrary, heart rate showed an increasing trend with walking time reaching a significantly higher maximum rate during placebo, with a difference of 5.4 bpm (2.9-8.7, p < 0.005) compared to oxygen. The distance walked was slightly but significantly greater after oxygen by 28 m (2-53, p < 0.05) and end of test dyspnoea after placebo by 0.6 points (0.1-1.1, p < 0.05). Respiratory rate increased over time, without differences between oxygen and placebo in the first minute of walking, then increasing significantly more during placebo (p < 0.0005). With placebo, tidal volume increased rapidly reaching a plateau at about 48% of FVC after 3 min, while with oxygen, the increase was slower, reaching a maximum of about 45% of FVC at the end of the test. Nevertheless, the difference was highly significant (p < 0.0005) at all the time points. Minute ventilation also increased significantly with walking time but remained at a highly significant lower level during oxygen than placebo at all the time points. Mean reduction in VE during the test with oxygen compared to placebo was 4.4 L/min (3.9-4.9, p < 0.0005). In our ILD patients, administration of outpatient oxygen during walking was related to a reduced increase in heart rate, respiratory rate, tidal volume and minute ventilation necessary to meet increased oxygen requirements, resulting in a lower workload on the cardiovascular system and on respiratory muscles and a consequent reduction in dyspnoea.