Sixty consecutive patients referred for evaluation of non-cardiac chest pain had oesophageal manometry. Motility was assessed basally, after edrophonium 80 micrograms/kg iv and during oesophageal perfusion with 0.1 N HCl at 6 and 14 ml/min for eight and seven minutes respectively. A positive response, defined as symptom reproduction with or without abnormal motility, was present in 21 patients (35%) after acid perfusion and 12 (20%) after edrophonium. Eleven of the 12 patients responding to edrophonium also responded to acid perfusion, including most of the patients with primary motility disorders. Significantly greater increases in peristaltic duration, but not amplitude, were recorded after edrophonium (p less than 0.01) and acid perfusion (p less than 0.05) in positive responders, compared with non-responders. Results indicate that acid perfusion during oesophageal manometry may be a more useful stress test than edrophonium and that the mechanism of symptom production may be similar.