Irwin and co-workers have designed an anatomic approach to the diagnosis and treatment of cough. In their hands, diagnosis was consistently determined and treatment successful almost without exception, if sustained. We reviewed the results of a similar approach in 139 consecutive and unselected patients referred to pulmonary specialists in two community hospitals. Thirty-nine patients demonstrated hyperreactive airways (HA) by carbachol inhalation and/or eucapnic hyperventilation of cold air. Twenty-seven of 78 without HA had postnasal drip, and 13 of 78 had a persistent cough following acute upper airway inflammation. Other less common diagnoses included chronic bronchitis, gastro-esophageal reflux, occupational bronchitis, interstitial lung disease, and psychologic causes. We were able to find the cause of cough 88 percent of the time. Treatment adjusted for noncompliance was not always a success. While all patients with HA improved, 8 percent of patients without HA or specific diagnosis did not have an improvement in their cough upon retrospective inquiry. Based on this analysis, we find that the diagnosis and treatment of cough may not be as successful as originally reported using Irwin's approach.