PGV performed in 39 patients by separating the lesser omentum from the stomach beginning 6 or 7 cm proximal to the pylorus and skeletonizing the distal 1 to 2 cm of esophagus was followed by 15.4% of proven and 10.2 of suspected recurrent ulcers. Insulin tests were done during the first 3 months postoperatively on 31 of the patients, including the 6 with proven and the 4 with suspected recurrent ulcers. The peak acid output to insulin minus tha basal acid output (PAOI-BAO) was less than 5 mEq/hr in 16 cases (52%) and from 5 to 25 mEq/hr in the remaining 15 cases. In 6 patients with proven recurrent ulcer, PAOI-BAO averaged 21.9 mEq/hr (range, 11.3 to 41.8); in the 4 patients with suspected recurrence, 9.5 (range, 4.4 to 11.8). The operative technique was changed in one respect; the distal 5 to 7.5 cm of the esophagus was skeletonized. In 14 patients, the mean PAOI-BAO +/- S.E. within 3 months of PGV was 1985 +/- 0.7 mEq/hr, and 13 of 14 values were less than 5 mEq/hr. One patient developed recurrent ulcer and required re-operation; this patient's value for PAO-BAO was 1.8 mEq/hr. The results show quantitatively that great differences in the completeness of PGV result from differences in the periesophageal dissection and emphasize its importance if optimal results are to be obtained and, especially, if the efficacy of the operation is to be judged.