Numerous studies show that asthma care is inadequate. Health insurance reimbursement data could prove useful for asthma surveillance. However, unlike the data available in other countries, diagnoses made during outpatient visits are not included. Using data from the 2006 Health and Social Protection Survey (ESPS), the quality of indicators based on drug reimbursements was examined to propose indicators for monitoring asthma care. The ESPS survey is conducted every two years among a sample of households in which at least one member is covered by social insurance. Matching with health insurance data from the current year allows for the retrieval of medical consumption data. In 2006, the self-administered health questionnaire included a module on asthma, and an additional matching with reimbursement data from 2004 and 2005 was performed for respondents enrolled in the general health insurance scheme. The validity of various algorithms based on drug reimbursement data for identifying asthma (or persistent asthma) was studied by calculating sensitivity, specificity, and positive (PPV) and negative (NPV) predictive values based on data reported by the respondents. Correlations between reimbursements for short-acting inhaled bronchodilators (SAB) and asthma control were also examined. It is difficult to define an indicator based on medication use that has both good PPV and good sensitivity for persistent asthma and, consequently, to propose indicators for monitoring the quality of care for persistent asthma. It is therefore proposed to monitor, using the general practitioner sample of beneficiaries (EGB) of the health insurance system, two populations corresponding to two situations of uncontrolled asthma: - heavy users of SBRAs and heavy users of inhaled corticosteroids (ICS) (difficult-to-treat asthma); - heavy users of BDCD and non-users, or light users, of IC (inadequately treated asthma). (R.A.)