OBJECTIVES. To analyse and evaluate current practice of asthma therapy by use of computerized prescription data. DESIGN. Descriptive, non-interventive. SETTING. A pharmacoepidemiological database on computerized refund claims from pharmacies in the Odense region in Denmark. SUBJECTS. Persons aged 20-44, presenting prescriptions for asthma drugs during the period 1 October 1990 to 30 September 1991. MAIN OUTCOME MEASURES. Prevalence of asthma drug use, the regimes chosen, relationship between beta-agonist and steroid use by the individual patient. RESULTS. In all, 3020 persons used anti-asthma drugs, corresponding to a crude 1-year prevalence of 3.6%. Sixty per cent were women. The subjects had an average anti-asthma drug consumption corresponding to 86% of the population's average. The four most common regimes were inhaled beta-agonist monotherapy, inhaled beta-agonist and corticosteroid, oral beta-agonist monotherapy and inhaled corticosteroid monotherapy, altogether accounting for 79% of users. The proportion of corticosteroid users increased with increasing individual use of beta-agonists. With an annual beta-agonist purchase of 100-200 defined daily doses, still less than half also used steroids. CONCLUSION. The majority of patients use a regime agreeing with current guidelines, except that more patients should be prescribed corticosteroids. The observed large increase in asthma drug use is likely to be explained by an intensified treatment of individual patients, rather than an increased prevalence.