The Bernstein test has been used as a test of oesophageal acid sensitivity for over 30 years but its clinical value has been challenged by the advent of ambulatory pH monitoring. Furthermore, the relation between mucosal acid sensitivity, symptomatic reflux, and abnormal oesophageal acid exposure time is unclear. This study examined the relation between these three parameters in patients referred for pH monitoring with unexplained chest pain or heartburn. Fifty consecutive patients were studied - nine with non-cardiac chest pain and 41 with a history of heartburn. Symptomatic reflux was defined as a greater than or equal to 50% temporal association between pain episodes and reflux events (pH less than 4) during pH monitoring. A positive acid perfusion test (in which the patient's usual symptoms were evoked by acid, though not saline) had a 100% sensitivity, 73% specificity, and 81% accuracy for the detection of symptomatic reflux. All 10 patients with symptomatic reflux during pH monitoring had evidence of mucosal acid sensitivity. A negative acid perfusion test made symptomatic reflux unlikely. However, symptomatic reflux or a positive acid perfusion test, or both, were found in some patients with a normal oesophageal acid exposure time during pH monitoring. Mucosal acid sensitivity, abnormal oesophageal acid exposure time, and symptomatic reflux should be regarded as separate, though related aspects of reflux disease. The Bernstein test is simple, safe, and easily performed. A positive test helps to identify an oesophageal cause of symptoms, particularly in patients in whom other aspects of 'gastro-oesophageal reflux disease' are absent, or who do not have symptoms during pH monitoring.