The authors modified the operative procedures used in pouch and interposition (PI) reconstruction in an attempt to improve the surgical results after total gastrectomy, because a randomized controlled trial had revealed that the clinical assessment of PI was quite poor, even though it is a physiological route. In most of the treated patients, the gastric emptying test revealed delayed emptying, and an X-ray video film showed folding and twisting of the jejunal conduit between the pouch and duodenum, which disturbed the transmission of nutrition. Modified PI (m-PI) was performed by decreasing the length of the jejunal conduit and widening the mesenteric pedicle to preserve the blood and nerve supply. This procedure was retrospectively compared with the previously used PI reconstruction by evaluating the postprandial symptoms, food intake, body weight, serum nutritional parameters, and emptying time of the gastric substitute. The m-PI group (n = 6) showed a lower incidence of symptoms, a greater food intake, and a greater weight recovery than the PI group (n = 6). The gastric emptying test also revealed an acceptable degree of emptying. We thus conclude that the m-PI reconstruction is more useful for improving the postoperative quality of life than the previously used method of PI reconstruction.