Respiratory failure can be caused by a variety of diseases and can lead to death due to poor oxygenation. The role of non-invasive respiratory support has been debated. This meta-analysis assesses the effectiveness and reliability of HFNC, standard oxygen (SO) treatment and non-invasive ventilation (NIV) in patients with respiratory failure respectively. The PubMed, Cochrane library, and CNKI databases were systematically searched from the inception dates to March 1, 2022. The primary randomized clinical trials included in meta-analyses were identified. The participants were patients with acute respiratory failure. Hospitalization mortality was defined as the primary outcome. Secondary outcomes were Failure of ventilation, Infection. The PROSPERO database has been registered with this meta-analysis. (registration number: CRD42022320088, 03/26/2022). A total of 26 RCTs involving 6518 patients were included. HFNC did not differ from NIV or SO therapy in terms of hospitalization mortality, ventilation failure, or lung infection. Patients with acute respiratory failure treaded with HFNC were more likely to develop remaining organ failure during hospitalization than those treated with NIV (P = 0.002, I 2 = 0%). Compared to SO, the use of HFNC leads to a more comfortable experience to patients (P=0.0003, I 2 =0%) and increase the oxygen partial pressure (P=0.001, I 2 =0%). In the subgroup analysis results of COVID-19, there were no significant differences between the HFNC, SO, and NIV for intervention failure, hospital mortality and oxygenation index. In hemodynamically stable patients with acute respiratory failure, there was no significant difference in in-hospital mortality and intervention failure rates between HFNC and SO and NIV. HFNC was superior to SO in improving patient oxygen partial pressure and comfort. In addition, there was no significant efficacy difference between NIV and SO for HFNC in the treatment of acute respiratory failure in COVID-19 patients.