The classification of asthmatics into severity categories is a crucial issue for assessing the asthma burden within a community, in which a proportion of patients is currently treated. There is no epidemiological method currently available.The Global Initiative for Asthma (GINA) was used to classify 4,362 patients aged 16±45 yrs (49% males, 42% taking inhaled corticosteroids), enrolled by 545 chest specialists in France with short standardized questionnaires including forced expiratory volume in one second (FEV1) measurements. Two independent GINA classifications were combined, one based only on symptoms and FEV1, and the other based only on current medication, to construct a final "symptom-FEV1 medication" classification.Almost 40% of the patients classed as step 1, 30% of those classed as step 2 and 13% of those classed as step 3 in the initial symptom-FEV1-classification, were allocated to categories of higher severity in the final classification. The approach was validated by showing that the proportions of: 1) patients considered by the physicians as having severe or moderately severe asthma; 2) patients with a history of hospital admission for asthma; and 3) patients with a history of emergency department visits for asthma, increased with severity steps in the final classification, for each step of the two initial independent classifications.The treatment manage plan in the Global Initiative for Asthma was not developed for assessing severity of asthma but rather to describe the recommended therapy for asthma with different severity. This is the first attempt to assess the severity of asthma in a large population of asthmatics mostly taking treatment, based on the Global Initiative for Asthma guidelines. The authors propose this simple and pragmatic procedure for a potential classification which should be put to the test in other studies. Eur Respir J 2000; 16: 615±620.