The symptom index is a quantitative measure developed for assessing the relationship between gastroesophageal reflux and symptoms. Controversy exists, however, over its accuracy and the appropriate threshold for defining acid-related symptoms of heartburn and chest pain. Therefore, a retrospective review was done of 153 consecutive patients referred to our esophageal laboratory. Three groups were identified: patients with normal 24-hr pH tests and no esophagitis, patients with abnormal 24-hr pH tests and no esophagitis, and patients with abnormal 24 hr pH values and endoscopic esophagitis. If symptoms occurred during the pH study, a symptom index (number of acid related symptoms/total number of symptoms x 100%) was calculated separately for heartburn and chest pain. Heartburn and chest pain episodes were similar among the three groups. However, the mean symptom index for heartburn was significantly (P < 0.001) higher in the patient groups with abnormal pH values [abnormal pH/no esophagitis: 70 +/- 7.1% (+/- SE); abnormal pH/esophagitis: 85 +/- 4.6%] as compared to those with normal studies, ie, functional heartburn (26 +/- 10.7%). The mean symptom index for chest pain was similar for all three groups. Using receiver operating characteristic curves, a heartburn symptom index > or = 50% had excellent sensitivity (93%) and good specificity (71%) for acid reflux disease, especially if patients complain of multiple episodes of heartburn. In contrast, an optimal symptom index threshold for defining acid-related chest pain episodes could not be defined.