Recently a consensus has emerged that health care research should address outcomes important to patients, especially quality of life, role performance, and functional status. The assessment of such outcomes is beset by conceptual and methodological difficulties that may be especially problematic for asthma. Nevertheless, several broad conclusions may be drawn about the use of measures of these outcomes in asthma research. Asthma usually is reasonably well controlled if patients are moderately adherent to their recommended regimens. Consequently, the beneficial impacts of interventions are likely to be small, and large samples are required to detect them. Outcome assessment should combine asthma-specific measures with generic measures applicable to a variety of conditions. Generic measures aimed at severely debilitating disease are less appropriate than measures designed for use in the general population. Asthma-specific measures should emphasize the incidence and impact of such symptoms as coughing, wheezing, sputum production, and shortness of breath. Current procedures for computing utility scores and cost-benefit ratios based on them have serious measurement limitations, and use of such scores should be postponed until those limitations are overcome. These assessment issues should be addressed separately for adults and children.