A prospective study was performed between June 1996 and December 1997, to identify how general practitioners (GPs) in Belgium assess asthma severity and how they treat asthma according to their severity assessment. Three hundred and sixty-five GPs included 1376 already diagnosed and treated asthmatics.The GPs used a questionnaire providing data on patient demographics, aetiology of asthma, symptoms and medication use. The patients provided a complete diary card of day and night symptoms and morning and evening peak expiratory flow rates during a 3-week period. Asthma severity as assessed by GPs was compared with the severity according to the GINA guidelines. Along the same line, asthma treatment was evaluated according to the GPs assessment of severity and according to the GINA guidelines. Confronting the assessment of asthma severity by the GPs with the GINA criteria revealed that about 20% and 2% of the patients' asthma severity respectively were under- and over-estimated, respectively (using a discrepancy between GPs and GINA assessment of severity by 2 or more classes). Using the GINA criteria for treatment, only 37.5% of the patients seemed to be correctly treated. Taking a discrepancy between GINA assessment and treatment of two classes as an error, 2.3% and 23.4% of the asthmatics are over- and under-treated, respectively. In conclusion, this study provides evidence that GINA guidelines seem not to be adequately interpreted and implemented by GPs in Belgium. Improvement of the assessment of asthma severity is definitely needed and may lead to more appropriate use of asthma medication.