The aim of this study was to determine whether we could measure exhaled nitric oxide (NO) levels in children, and whether the same pattern of exhaled NO concentrations was observed in asthmatic and normal children as had been seen in adults. Using a chemiluminescence NO analyzer, we measured NO in exhaled air both directly and through a T-piece allowing us to measure carbon dioxide (CO2), mouth pressure, and expiratory flows. In 39 normal children the mean peak exhaled NO was 49.6 parts per billion (ppb) (SD 37.4) when all expired gas passed directly through the NO analyzer, and 29.7 ppb (SD 27.1) when expiration occurred through a T-piece. The results were significantly higher in 15 asthmatic subjects on bronchodilator therapy only [126.1 ppb (SD 77.1) direct (P < 0.001), and 109.5 ppb (SD 106.8) via T-piece (P < 0.001)]. In 16 asthmatics on regular inhaled corticosteroids the mean peak exhaled levels were significantly lower 48.7 ppb (SD 43.3) direct (P < 0.001) and 45.2 ppb (SD 45.9) via T-piece (P < 0.01). There was no difference between the normal children and the asthmatic children on regular inhaled corticosteroids (P = 0.9 direct, P = 0.2 via T-piece). There were no significant differences in carbon dioxide levels, mouth pressure, duration of expiration and expiratory flows between the different groups, and no difference between carbon dioxide levels, mouth pressure and duration of expiration between the two methods (direct and T-piece). In 6 asthmatic children mean peak exhaled levels on NO fell from a median peak level of 124.5 ppb to 48.6 ppb when measured before and 2 weeks after commencement of inhaled corticosteroid treatment. The measurement of exhaled NO levels may be useful as a noninvasive means of monitoring children with asthma.