AbstractWe aimed at evaluating pressure transmission and stability during non‐synchronized neonatal nasal intermittent positive pressure ventilation (NIPPV) delivered using five mechanical ventilators and three nasal interfaces. An artificial nose–throat model was connected to a mechanical analog of the infant respiratory system and a breath generator. Ventilation was administrated via a nasal mask (NM), short bi‐nasal prongs (SBN), or RAM® cannula. We applied positive end‐expiratory pressures (PEEP) of 5 and 10 cmH2O, inspiratory pressures (PIP) of 15 and 30 cmH2O, inspiratory times of 0.23, 0.42, and 0.57 s. Measurements were performed with leaks of 0, 1.5, and 4 L/min. The pressure was measured at the airways opening (PAW) and the glottis (PGL). The difference between set and delivered pressures (PAW) was less than ±1 cmH2O for all ventilators. We documented a significant difference between PAW and PGL in the presence of leaks. With 4 L/min leaks, PEEP dropped by 43%, 49%, and 63% with NM, SBP, and RAM® cannula, respectively; PIP dropped by 58%, 64%, and 74%. On average, the SD of PEEP fluctuations was ±0.60 and ±2.50 cmH2O for PAW and PGL; the breath‐by‐breath SD of PIP was ±0.77 and ±2.06 cmH2O. During NIPPV, the PIP and PEEP transmission to the glottis is markedly lower than the set values and highly variable. The impact of leaks and nasal interface is much more significant than the differences in ventilators' performance on the efficacy of pressure transmission and stability of non‐synchronized ventilator‐generated NIPPV.