Nine patients complaining only of chronic cough of unknown cause were prospectively studied with prolonged esophageal pH monitoring (EPM) before and after cough had disappeared as a complaint in order to determine if and why gastroesophageal reflux (GER) was causing their coughs. Coughs disappeared as a complaint an average of 161 +/- 75 days after medical therapy for GER. Comparisons of pretreatment and post-treatment EPM data revealed the following: numbers of coughs (p = 0.029), total refluxes (p = 0.001), refluxes greater than or equal to 5 min (p = 0.019), and reflux-induced coughs (p = 0.005) had significantly decreased in the distal esophagus, and total refluxes (p = 0.05) had significantly decreased in the proximal esophagus. During the entire study period, the number of coughs were significantly correlated with the number of total refluxes (p = 0.039), longest reflux (p = 0.019), number of refluxes greater than or equal to 5 min (p = 0.006), and percent of total time that pH was less than 4 (p = 0.017) in the distal esophagus. On the basis of these results, we conclude that (1) cough can be the sole presenting manifestation of GER, and it gradually responds to standard GER therapy; (2) prolonged EPM is safe, well-tolerated, and extremely useful in diagnosing clinically silent GER; (3) the mechanism by which GER causes cough is related to a critical number and/or duration of reflux episodes in the distal and/or proximal esophagus.