Over the past decade and boosted by the coronavirus disease 2019 (COVID-19) pandemic, high-flow nasal oxygen (HFNO) has been increasingly used in the intensive care unit (ICU) to treat acute hypoxemic respiratory failure (AHRF). In this review, we show that despite this wide and rapid increase in the use of HFNO to treat AHRF, HFNO does not fulfill all the criteria of a "reference treatment". First, there are some inconsistencies between the studies that provided a positive signal toward the possible benefit of HFNO in AHRF. The two high-quality studies were negative in terms of primary outcome although they provided promising signals in favor of HFNO in terms of secondary outcomes or unplanned secondary analysis. The significance of the only positive study suffers from notable limitations and other trials, conducted in COVID-19 and in immunocompromised patients, are definitely negative and do not even provide promising signals in favor of HFNO. Of note, authors of some of the large randomized controlled trials (RCTs) on HFNO have received grants or personal fees from manufacturers of HFNO devices. Second, meta-analyses do not show positive results regarding the efficacy of HFNO on mortality and recent guidelines do not support its use to improve this outcome, although they recommend HFNO use to reduce intubation rate. Third, HFNO is associated with risks that should be accounted for. There are concerns that HFNO may delay intubation, which is in turn associated with higher mortality and prolonged length of stay. In addition, with HFNO, high inspiratory effort may generate high lung strain and overstretch, a phenomenon termed patient self-inflicted lung injury (P-SILI). Fourth, there are concerns regarding access to HFNO in resource-limited settings. Fifth, there are also concerns regarding the deleterious environmental impact of HFNO due to the high volume of consumables and high oxygen flow, which remain to be precisely quantified and balanced with the potential reduction in intubation rate. Considering all these limitations, HFNO is not yet the reference treatment for AHRF.