Although several studies have examined early outcome following redo antireflux surgery, there is little data on the long-term efficacy of these procedures. We reviewed our experience of these operations in order to assess the long-term results which can be achieved by choosing redo antireflux procedures based on the results obtained from pre-operative oesophageal function testing. The case notes of 26 patients who underwent a repeat antireflux procedure between 1981 and 2000 were reviewed. Clinical history, examination, endoscopy, pH studies, oesophageal manometry and video barium contrast studies were performed on all patients prior to re-operation. In all cases, a standardised pre-formulated algorithm, based on the results of the pre-operative oesophageal function tests, was used to determine the choice of surgical procedure. Post-operatively, patients were classified into three groups: Group A (those with no symptoms), Group B (symptoms controlled by medication) and Group C (symptoms refractory to medical treatment). Patients in Groups B and C underwent repeat oesophageal function studies. In all cases, the indication for re-operation was gastro-oesophageal reflux symptoms refractory to medical treatment. Twenty-one patients (81%) underwent a left thoracotomy, of whom 11 patients (42%) underwent a gastroplasty. The mean follow-up period was 8.27 years (range 1.5-19.8 years), after which 14 patients (54%) were classified as Group A; 10 patients (38%) as Group B; and 2 patients (8%) as Group C. Within Groups B and C, manometry showed that re-operation had increased basal lower oesophageal sphincter pressure (4.6 vs 12.7 mmHg, pre- vs post-operative P=0.03), and in all, but one case pH studies showed no evidence of recurrent acid reflux. Redo antireflux surgery can provide complete symptomatic relief in approximately 50% of patients and symptomatic improvement over 90% of patients. We advocate a tailored approach in the selection of re-operative procedures based on the results of pre-operative oesophageal function testing.