Neonates with chronic lung disease often require oxygen in the neonatal intensive care unit. The purpose of this study was to determine (1) the actual inspired oxygen concentration (FiO2) delivered to neonates when using a low-flow flowmeter and a nasal cannula, and (2) the accuracy with which FiO2 could be estimated using a formula that we developed. We studied two groups of infants: 18 infants less than 1,500 g and 13 infants greater than 1,500 g. We measured pharyngeal oxygen levels by sampling pharyngeal gas in infants receiving 100% humidified oxygen by nasal cannula from a low range flow flowmeter. The oxygen flow was increased by 25 mL/min increments from 25 to 200 mL/min. The measured FiO2 was compared with the calculated FiO2 using the formula: [formula: see text] where minute ventilation (VE) equals the minute ventilation in mL/min (VE = VT X respiratory rate). For both groups of infants, increments of 25 mL/min of flow produced distinctive changes in FiO2 at all levels (P < 0.001). The calculated FiO2 did not significantly differ from the actual FiO2 at any flow. The calculated FiO2 was most predictive when using an assumed tidal volume of 5.5 mL/kg. We conclude that an accurate flowmeter connected to 100% humidified oxygen can produce a wide range of predictable FiO2s for neonates, especially those with birthweights of less than 1,500 g. The proposed formula allows useful estimation of the infant's FiO2 when we assume a tidal volume of 5.5 mL/kg.