Measurement of the fraction of exhaled nitric oxide (FENO) has been proposed as a noninvasive assessment of asthmatic airway inflammation. The influence of the expiratory flow rate during the collection maneuver on the ability of FENO to discriminate healthy subjects from those with asthma is unknown. We compared online and offline measurement of FENO at different flow rates. FENO was collected with expiratory flows of 50-500 ml/second in 34 patients with asthma (PC(20) of less than 8 mg/ml) and 28 healthy subjects (PC(20) of more than 10 mg/ml) using offline collection techniques. In a subgroup of 18 individuals with asthma and 17 healthy subjects, we additionally measured FENO at multiple expiratory flow rates (47-250 ml/second) using online methods. FENO fell with an increasing expiratory flow rate; FENO was higher in subjects with asthma as compared with healthy subjects at each flow rate studied with both techniques (p < 0.001). Receiver operating characteristic (ROC) curves for the diagnosis of asthma indicated that FENO is a robust discriminator between individuals with asthma and healthy subjects (area under the ROC curves 0.79 +/- 0.06 to 0.86 +/- 0.06, p for significant discrimination < 0.0001). Neither expiratory flow rate nor collection technique (online versus offline) significantly altered this discriminatory capacity (area under the ROC curves = 0.84 +/- 0.07 with the slowest online method versus 0.80 +/- 0.07 with the fastest offline method, p = 0.46). These data indicate that the choice of expiratory flow rate and collection method can be based on practicality and patient comfort without compromising the utility of this test for asthma.