Background Although there are current measures to evaluate childhood asthma severity for clinical diagnosis and treatment, there is no standard valid measure to evaluate childhood asthma severity for large-scale epidemiologic studies. Objectives To develop and test a childhood asthma severity scale (CHAS) for clinimetric validity and to determine differences in symptoms, medication use, and health care visits by participant characteristics. Methods Eight hundred ninety-seven actively asthmatic children under the age of 12 years were selected from a general population of children. Children were selected from a screening questionnaire administered at six Connecticut hospitals that serve large minority populations in Bridgeport, New Haven, Hartford, and Danbury and one hospital serving south central Massachusetts. Twelve-month baseline data for a prospective cohort study of childhood asthma severity were collected on a monthly basis through home interviews. Home interviews addressed questions on daily symptoms, medication use, and health care visits. A severity scale was constructed using three dimensions: symptoms, medication use, and health care visits. Results CHAS has sufficient preliminary content, construct, and predictive validity. Despite similarities in symptoms, there were health care utilization and medication differentials according to race and ethnicity, insurance status, family income, and maternal education. Conclusions CHAS is a potentially useful measure of asthma severity for large-scale epidemiologic studies. It seems that CHAS has sufficient clinimetric properties.