Objectives: Auto-inhalation of nitric oxide (NO) produced in the upper airways may have physiologic effects on lung function. For intubated patients, the upper airway source of NO is eliminated, but the hospital compressed air source from the environment is contaminated with varying levels of NO, creating an “occult” form of NO therapy. We examined the physiologic significance of occult inhaled NO in ventilator-dependent pediatric patients. We hypothesized that very low levels of NO contamination in inspired gas improve Pao₂ in ventilator-dependent children. Study design: Inspired NO levels at the mouth were measured by chemiluminescence in 4 pediatric subjects with normal lungs and 3 with parenchymal lung disease. Subjects were sequentially ventilated with first standard hospital gas (H1), switched to pure nitrogen-oxygen at a similar Fio₂ but with no NO contamination (A2), hospital gas again (H2), the nitrogen-oxygen (A2) to control for time and sequence, and finally the nitrogen-oxygen mixture with supplemental NO in an amount equal to the NO previously measured in hospital gas (A2 + NO). Inhaled NO levels and Pao₂ were recorded 15 minutes into each of the 5 steps. Two patients were studied a second time, remote from their first examination. Results: NO levels in inhaled hospital gas mixtures ranged from 13 to 79 ppb (mean H1 = 53.3 ± 23.7 ppb, mean H2 = 53.2 ± 20.7 ppb, mean A2 + NO = 45 ± 15.3 ppb; P < .0001). Removing NO from ventilator gas decreased Pao₂ in all subjects, whereas replacing NO in artificial gas restored Pao₂ to baseline values (P < .0001). Conclusion: Concentrations of NO in hospital compressed air are variable and have physiologic effects. The long-term implications of these findings remain to be defined. (J Pediatr 1998;133:613-6)