The objective was to determine whether high flow nasal cannula (HFNC), a promising respiratory support in infant bronchiolitis, could reduce the proportion of treatment failure requiring escalation of care.In this RCT, we assigned infants younger than 6 months who had moderate bronchiolitis to receive either HFNC at 3 L·kg-1·min-1 or standard oxygen therapy. Cross-over was not allowed. The primary outcome was the proportion of patient in treatment failure requiring escalation of care (mostly non-invasive ventilation) within 7-days following randomisation. Secondary outcome included rates of transfer in PICU, oxygen and nutritional-support length of days and adverse events.The analyses included 268 patients among the 2621 infants assessed for inclusion during 2 consecutive seasons in 17 French pediatric emergency departments. The percentage of infants in treatment failure was 14% (19 of 133) in the study group, compared to 20% (27 of 135) in the control group (OR 0.66; CI 95% 0.35-1.26, p=0.21). HFNC did not reduce the risk of admission in PICU (21 of 133 in the study group (15%) versus 26 of 135 in the control group (19%))(OR 0.78; CI 95% 0.41-1.41, p=.45). The main reason for treatment failure was the worsening of modified-Wood clinical asthma score (m-WCAS). Short-term assessment of respiratory status showed a significant difference for m-WCAS and respiratory rate in favor HFNC. Three pneumothoraces were reported in the study group.In patients with moderate bronchiolitis, there was no evidence of lower rate of escalating respiratory support among those receiving HFNC therapy.