The aim of this study was to compare exhaled nitric oxide concentrations obtained during controlled slow exhalation, presently considered as the method of choice, with two sampling methods that are easily performed by children: blowing air into a balloon and tidal breathing through a mouthpiece. One hundred and one well controlled, stable allergic asthmatic children (median age 11.7 yrs) performed the following tasks in duplicate: 1) exhalation from total lung capacity through a mouthpiece against a resistor with a standardized flow rate of 20% of the subject's vital capacity per second, using a biofeedback system; 2) a single deep exhalation into an NO-impermeable mylar balloon; and 3) tidal breathing through a low resistance mouthpiece over 2 min. NO was measured using a chemiluminescence analyser. Twenty-nine children (29%) were not able to perform a constant-flow exhalation of at least 3 s. All children performed the balloon and tidal breathing methods without difficulty. NO concentrations (means +/-SEM) were 5.3+/-0.2 parts per billion (ppb) at the end-expiratory plateau, 5.2+/-0.3 ppb in balloons (intraclass correlation coefficient (r(i)) = 0.73) and 8.0+/-0.4 ppb during tidal breathing (p<0.001, r(i) = 0.53 compared to plateau values). Mean values of NO during tidal breathing increased significantly with time, suggesting increasing contamination with nasal air. It was concluded that, in asthmatic children, the end-expiratory plateau concentration of nitric oxide during exhalation at 20% of the vital capacity per second is similar to the values obtained with the balloon method, with satisfactory agreement, but differs from values obtained during tidal breathing. The balloon method is cheap, simple and offers the interesting possibility to study exhaled nitric oxide in young children independently of the presence of a nitric oxide analyser.