The COVID-19 pandemic has highlighted global oxygen system deficiencies and revealed gaps in how we understand and measure oxygen access . We present a case study on oxygen access from 58 health facilities in Lagos state, Nigeria . We found large differences in oxygen access between facilities (primary vs secondary, government vs private) and describe four key domains to consider when measuring oxygen access . Use : 8/58 (14 %) of facilities had a functional pulse oximeter for detecting hypoxaemia (low blood oxygen level) and guiding oxygen care . Oximeters were typically located in outpatient clinics (12/27 , 44 %), paediatric ward (6/27 , 22 %), or operating theatre (4/27 , 15 %), not suitable for children, and infrequently used . Availability : 34/58 (59 %) facilities had a functional source of oxygen available on the day of inspection, of which 31 (91 %) facilities had it available in a single ward area, typically the operating theatre or maternity ward . Cost: Oxygen was free to patients at primary health centres, when available, but expensive in hospitals and private facilities, with the median cost for 2 days oxygen 13000 ($ 36 USD) and 27500 ($ 77 USD) naira, respectively . Patient access: No facilities were adequately equipped to meet minimum oxygen demands for patients . We were unable to determine the proportion of hypoxaemic patients who received oxygen therapy with available data . We highlight the importance of a multi-faceted approach to measuring oxygen access that assesses access at the point-of-care, and ideally at the patient-level . We propose standard metrics to report oxygen access and describe how these can be integrated into routine health information systems and existing health facility assessment tools.