The purpose of this study was to evaluate the effectiveness of trans-esophageal atrial stimulation in decreasing atrial flutters. 31 patients, aged between 26 and 86 years, underwent 38 esophageal stimulations between August 1986 and April 1987. Esophageal stimulation was carried out with a device delivering major stimuli (10 to 20 ms) as well as high voltages (10 to 20 volts). A bipolar probe of permanent intracardiac stimulation was placed behind the left atrium, via a trans-esophageal approach under ECG guidance. Stimulations were carried out at a slightly faster rhythm than that of the atrial flutter, for 1 to 30 seconds with a progressive increase of the stimulation frequency (280 to 960 impulses/min) until either a sinus rhythm or an atrial fibrillation was obtained. A sinus rhythm was obtained immediately (21 times out of 38; 55.3%) and at the 24th hour after temporary atrial fibrillation in 8 additional patients (21%). The reduction percentage at 24 hours was therefore 76.3 p. cent. In 6 patients (15.8%), esophageal stimulation failed and in 3 additional patients, after atrial fibrillation, there was either relapse into flutter, or persistence of atrial fibrillation at the 24th hour. Esophageal stimulation was well tolerated in all cases. There were no local or rhythm complications. In conclusion, trans-esophageal atrial stimulation appears to be effective, easy to implement, economical and well tolerated. Its primary use to decrease atrial flutter seems justified, permitting to avoid in 75 p. cent of the time, recourse to intracardiac atrial stimulation or external shock.