Domiciliary long-term oxygen therapy (LTOT) is usually supplied by means of oxygen concentrators (OCs). Various factors that determine the efficacy of such a treatment were evaluated. Sixty-three patients, arbitrarily selected from lists of health care providers, were visited at home by a biomedical engineer and a pulmonary function technician. The evaluation consisted of: i) responses to a directed questionnaire, ii) assessment of the OC output characteristics, and iii) measurement of the patient's oxygen saturation (SaO2) at rest with and without oxygen supplement. Only 33% of patients received oxygen treatment for the recommended 12-24 hours/day and 5% of patients waited the recommended 10 minutes of OC warm-up before connection. Filters were cleaned weekly by only 30% of patients and the concentrator was serviced 3-4 times a year in 25% of cases. The OC was thought to be unduly noisy by 24% of patients and connecting tubing of less than 6 meters was fitted to 90% of OCs, thereby limiting patient mobility. Most of the OCs did not yield the recommended oxygen concentration and the flow rate meters on them tended to underread. Therefore, only 22% of patients received the prescribed oxygen supplement. Whilst breathing room air, a substantial proportion of patients had an SaO2 >90%. Improvements are clearly required in terms of medical indications for LTOT, patient education and supervision, supply and maintenance of concentrators and related equipment.