Background: Cough is widely recognized as a key symptom in the diagnosis and the monitoring of asthma, but little is known about how best to assess cough in asthma. Objective: To determine how objective cough rates correlate with subjective measures of cough in asthma. Methods: We studied 56 subjects, median age 42.0 years (range, 28.5-71), 34 (60.7%) female, with asthma. Subjects performed cough reflex sensitivity testing (concentration of citric acid causing 2 and 5 coughs [C2 and C5]), 24-hour fully ambulatory cough recordings, subjectively scored the severity of their cough (visual analog scales and 0-5 score) and completed a cough-related quality of life questionnaire (Leicester Cough Questionnaire). Ambulatory cough recordings were manually counted and reported in cough seconds per hour (cs/h). Results: The median time spent coughing was 2.6 cs/h (range, 0.0-14.2), with subjects spending more time coughing by day (median, 3.9 cs/h [0.0-18.5]) than by night (median, 0.3 cs/h [0.0-8.7]; P <.001). A weak inverse relationship was seen between day cough rates and log10C2 (r = -0.39; P = .03) but not log10C5 (r = -0.08; P = .65). Objective time spent coughing was also weak-moderately associated with subjective cough scores and visual analog scales, and most strongly correlated with cough-related quality of life (r = -0.54; P <.001). Conclusion: Subjective measures of cough and cough reflex sensitivity are poor surrogates for objective cough frequency in asthma. When designing studies to assess interventions for cough in asthma, we advocate a combination of both objective measures of cough and cough-related quality of life. © 2008 American Academy of Allergy, Asthma & Immunology.