A prospective randomized study was done on 45 patients to evaluate the effectiveness of the Hill, Nissen or Belsey antireflux procedure. All patients had symptoms of GE reflux unresponsive to medical therapy, a + standard acid reflux test (SART), and esophagitis (38/45) or + Burnstein test (7/45). Esophageal symptomatic, radiographic, manometric and pH (SART and 24 hr monitoring) evaluation was done pre- and 154 days (average) postsurgery. All procedures improved the symptoms of pyrosis. The best improvement was seen after the Nissen repair. All procedures increased the distal esophageal sphincter (DES) pressures over preoperative levels. The Nissen and Belsey increased it more than the Hill. Sphincter length and dynamics remained unchanged. The Nissen procedure placed more of the manometric sphincter below the respiratory inversion point in the positive pressure environment of the abdomen. The esophageal length was increased by the Nissen and Hill repairs. This was thought to account for the high incidence of temporary postsurgery dysphagia following the Nissen and Hill repairs and the lower incidence following the Belsey repair. Reflux was most effectively prevented by the Nissen repair as shown by the SART and 24 hr esophageal pH monitoring, a sensitive measurement of frequency and duration of reflux. The average length of hospital stay was 20 days for Belsey and 12 days for both Nissen and Hill procedure. Postsurgery complications were more common following the thoracic than the abdominal approach. Ability to vomit postrepair was greatest with the Hill and least with the Belsey and Nissen repair. All procedures temporarily increased amount of flatus. It is concluded that the Nissen repair best controls reflux and its symptoms by providing the greatest increase in DES pressure and placing more of the sphincter in the positive abdominal enviroment. This is accomplished with the lowest morbidity but at the expense of temporary postoperative dysphagia and a 50% chance of being unable to vomit after the repair.