The effect of magnesium chloride infusion to different sites in the gastrointestinal tract of sheep upon the net absorption of magnesium from different regions of the digestive tract has been examined. Four Merino wethers were each prepared with cannulas sited in the rumen, in the duodenum adjacent to the pylorus and in the terminal ileum. The basal diet provided 18.45 mmoles magnesium/ day and an additional 65 mmoles magnesium/day as magnesium chloride was continuously infused into (A) the rumen, (B) the rumen and duodenum in equal portions, (C) the duodenum and (D) the terminal ileuni. A continuous infusion of Cr-EDTA to the ruinen enabled digcsta and magnesium flow rates to be estimated from digesta samples obtained from the intestinal cannulas at 4 hr intervals over 3 days. For treatments A, B, C and D respectively, the mean net absorption of magnesium (mmoles/day) from the rumen was 20.4, 11.4, 1.4 and 3.4; from the small intestine –0.5, 1.7, –5.1 and –9.8; from the large intestine 4.6, 2.2, 12.7 and 12.3; and from the total gastrointestinal tract 24.6, 15.4, 9.1 and 4.9. In each case the effect of treatment was significant. The total net absorption of magnesium caudal to the pylorus was unaffected by treatment. Plasma magnesium levels were reduced during post-ruminal infusion of magnesium, but these changes were not obviously linked to the changed net absorption from the intestinal segments. The urinary and faecal excretion of magnesium, but not the magnesium balance, was strongly related to the total net absorption of magnesium. The results emphasize the major contribution of the stomach to the gastrointestinal net absorption of magnesium and show that although the amount absorbed from this region may influence separately the net absorption from the large and small intestine, it does not appear to influence the overall intestinal net absorption of magnesium.