In order to define impairment from asthma the physician must resolve four issues: there must be a clear, unambiguous definition of disease; clear and precise criteria for disease severity must be established; the exposure must contribute to impairment; and the natural history or prognosis of the disease must be generally known. Each of these areas is discussed and existing data is reviewed to provide a framework for considering asthma impairment. Approaches for identifying persons with asthma include clinical evaluation, questionnaire history of diagnosis, questionnaire history of symptoms, response to bronchodilator, and bronchoconstrictor response. Epidemiologic measures of asthma severity include the level of lung function, degree of bronchial responsiveness, gender, race, atopy, and smoking habits. Clinical measures of asthma severity include the duration of disease, emergency room visits, hospitalizations, physician visits, circadian variability, medication use, and wheeze symptoms. A number of studies have indicated that allergy as measured by skin tests is associated with increased airways responsiveness. However, the odds ratios for skin test score and bronchial responsiveness vary from 0.6 to 3.5 in epidemiologic studies in adults. The authors conclude that a large number of factors thought to influence asthma severity have not been rigorously tested.