This study compared estimates of the severity and impact of asthma recorded using global questions of the type used in diary cards with health status measurements obtained using comprehensive questionnaires. Seventy-four outpatients with asthma, aged 17-76 yrs (mean 48 yrs) participated. Mean+/-SD forced expiratory volume in one second (FEV1) was 72+/-26% predicted. Patients recorded morning and evening peak expiratory flow rate (PEFR) and scaled their responses to the questions: "How has your asthma been today?" (asthma severity) and "How much effect has your asthma had on your life today?" (asthma impact) for 2 weeks. They then completed Juniper's Asthma Quality of Life Questionnaire (AQLQ) and the St George's Respiratory Questionnaire (SGRQ). Diary card scores for asthma impact were less severe than for asthma severity (p<0.0001). Both correlated with AQLQ and SGRQ total scores (r>0.7; p<0.0001). Some patients responded 'none' for asthma severity (n=10) or asthma impact (n=13) on all 14 days of recording. For these patients, FEV1 was <80% predicted, morning PEFR was <90% predicted and their AQLQ and SGRQ scores indicated significant health impairment. Diary card scores for asthma severity and impact were correlated with health status, but these global questions were insensitive in mild disease. Responses to these questions were influenced by their wording, so the number of symptom-free days calculated from diary cards will depend on the questions used. Standardization is required before symptom-free days can be used as a reliable measure of treatment efficacy.