A retrospective review was carried out of 148 consecutive personal truncal vagotomies and anterior pylorectomies (TV + P), median follow-up 5.0 years. The recurrent ulcer rate was 6/148 (4.1%) for suspected (SRU) and 5/148 (3.4%) for proven ones (RU). This led to one death from RU. Thirty-six patients (24.3%) developed postvagotomy diarrhoea (PVD). Two of these were graded Visick IV because their occupations made PVD particularly inconvenient. These results are similar to those for TV and pyloroplasty, despite the slightly more destructive nature of pylorectomy. There were nine patients in whom evidence of associated bowel disease had been documented before or during operation. The occurrence of such evidence was significantly more frequent in the PVD group (6/36 vs. 3/112 patients, P = 0.014), suggesting either a summation of effects due to the bowel disease and the operation or that the diagnosis of PVD was sometimes incorrect. Of the six in the PVD group, two were in Visick Grade II because of their PVD, and four in Visick grade III or IV, but in one of these SRU was the main cause of the poor result, and in two the PVD was subsidiary to vomiting or dumping. One further patient in Visick III due to PVD had diverticular disease diagnosed after operation. Thus there were identifiable factors (occupation, associated bowel disorder) which could have been used to predict seven of the unsatisfactory results due partly or completely to PVD. TV + P is an effective operation for duodenal ulcer but, as with TV + pyloroplasty (though to no greater an extent), severe diarrhoea may occasionally mar the result.(ABSTRACT TRUNCATED AT 250 WORDS)