In adults fatal and near-fatal asthma have similar clinical characteristics. Therefore, near-fatal asthma in adults can be used as a model for fatal asthma. A nationwide study on fatal and near-fatal asthma in children <16 yrs was performed in order to assess whether, as in adults, near-fatal asthma can be used as a model for fatal asthma. From 1996 to 1998, all paediatric hospitals and paediatric pulmonologists in Germany were asked to report cases of fatal asthma and near-fatal asthma to a central survey unit (Erhebungseinheit für seltene pädiatrische Erkrankungen in Deutschland (ESPED)) on a monthly basis. All reports were followed by detailed questionnaires. Sixteen fatal and 45 near-fatal asthma cases were analysed. Fatal asthma patients were older than near-fatal asthma patients. Respiratory tract infections were frequently reported only in near-fatal asthma (47 versus 0%). The proportion of cases with rapid-type onset (duration of symptoms < or =1 h) was higher in fatal asthma (53 versus 14%). Long-term regular treatment with short acting beta2-agonists was common in both groups, but the use of concomitant inhaled corticosteroids was significantly lower in fatal asthma cases. A high proportion of poor compliance was observed in both groups. As fatal and near-fatal asthma differ significantly in important clinical aspects, analysis of near-fatal asthma might be of limited value in elucidating the causes of fatal asthma in children.