Hypoxia is a common complication seen in people with COVID-19 and can often be the presenting symptom. Using a multi-centre observational database, we analysed 3,624 hospitalised COVID-19 PCR-positive patients at Methodist Health System, Dallas, Texas, USA from March 2020 to December 2020. We compared in-hospital death or hospice referral rates and major adverse cardiovascular events (MACE) between patients with four levels of oxygen (O2) requirements (0-1 L/min, 2-10 L/min, 11-20 L/min, 21-100 L/min). MACE included congestive heart failure (CHF) exacerbations, myocardial infarctions (MI), strokes, pulmonary embolism (PE) / deep venous thrombosis (DVT), and shock. Logistic regression analysis was used to determine comorbidities and demographics associated with mortality. Multinomial regression analysis was used to find which of these variables were associated with hypoxia. Patients who arrived needing 0-1 L/min of O2 had reduced risk of mortality compared to those requiring 2-10 L/min (OR=1.54, 95% CI=1.207-1.976, p<0.0001), 11-20 L/min (OR=4.55, 95% CI=3.169-6.547, p<0.0001), or 21-100 L/min (OR=12.06, 95% CI=8.548-17.016, p<0.0001). In addition, patients who arrived needing 0-1 L/min of O2 showed reduced risk of MACE compared to those requiring 2-10 L/min (OR=1.20, 95% CI=1.029-1.409, p<0.0001), 11-20 L/min (OR=2.76, 95% CI 2.06-3.696, p<0.0001), or 21-100 L/min (OR=6.74, 95% CI 4.966-9.155, p<0.0001). Hypoxia on arrival is associated with a significantly increased risk of mortality and MACE among hospitalised patients with COVID-19. This data will promote better prognostication and help reduce negative outcomes in an inpatient setting.