Background: Dyspnoea, or distressing breathing, is often a severe subjective symptom in terminal illness and may be difficult to control. Oxygen therapy is currently one of the interventions used to treat it. This review aimed to identify all randomised controlled studies (RCTs) in which oxygen therapy was used as a treatment to relieve dyspnoea in chronic terminal illness, and to synthesize the findings into a conclusion regarding the overall effectiveness of oxygen therapy for the palliation of dyspnoea in chronic terminal illness. Objectives: The objective of this review was to determine if oxygen therapy, administered in a non-acute care setting, provided additional relief of dyspnoea in study participants with chronic end-stage disease over that provided by breathing room air or placebo air as a control. Search methods: Electronic databases were searched using predefined search terms. Searches were current to April 2006. Selection criteria: Only RCTs were considered for inclusion in this review. Unblinded studies were included. Data collection and analysis: Data was extracted by one review author and checked by another. Main results: Eight studies met the inclusion criteria for the review and included a total of 144 participants (cancer; n = 97, cardiac failure; n = 35, kyphoscoliosis; n = 12). Four cross-over studies, two studies with the participants at rest and two involving exercise testing, compared oxygen inhalation to air inhalation for dyspnoea management in adults with advanced cancer. Three cross-over studies compared the use of oxygen inhalation to air inhalation in adults with stable chronic heart failure for dyspnoea management during exercise testing and one crossover study compared ambulatory oxygen therapy with air inhalation on exercise-induced dyspnoea for study participants with kyphoscoliosis (a sideways and forwards curvature of the spine). No studies with matched or cohort controls were identified. Due to differences in study designs, few studies could be pooled for a meta-analysis. This systematic review of the literature failed to demonstrate a consistent beneficial effect of oxygen inhalation over air inhalation for study participants with dyspnoea due to end-stage cancer or cardiac failure. Some cancer study participants appeared to feel better during oxygen inhalation. Authors' conclusions: The failure to demonstrate a beneficial effect for oxygen breathing over air breathing in cancer or cardiac failure was limited by the small volume of research studies available for inclusion, the small numbers of participants and by the methods used in the studies.