A double-lumen tube with a mercury weight attached to its distal end was positioned in the small intestine of six subjects who were to undergo an elective laparotomy. The tube and the technique for its localization were exactly similar to those used in previous intestinal perfusion experiments. Marked ;concertinaing', or gathering of the small intestine proximal to the mercury weight, was seen at laparotomy in all six subjects. The ratio between the mean intestinal length from the mercury weight to the ligament of Treitz after removal of the tube to that with the tube in situ was 3.0 (2.5-3.8). Mean total jejuno-ileal length was 421 (320-521) cm. In five of the six subjects the distal end of the tube was more than half way between the ligament of Treitz and the ileo-caecal valve although it was only 100-120 (mean 108) cm from the incisor teeth. Segmental perfusion studies of the human jejunum involve a much longer length of small intestine than is generally assumed-by a factor of approximately 3. Such studies assess absorption rates over a substantial proportion of the small intestine and are not confined to a short segment of proximal jejunum. The nutritional significance of such studies in which amino acids, peptides, and carbohydrates are investigated is increased. Studies designed to measure absorption rates from the ileum should be treated with caution as part or all of the perfusion segment may be past the ileo-caecal valve.