The effect of the soluble chlorides of Mg++ and of Ca++ on intestinal water absorption was compared in rodents and in humans because Mg++ is a laxative while Ca++ is reputed to be constipating. In rats, net water transport in vivo from segments of jejunum or of colon was similar during control periods with NaCl, 275 mosm/kg, and during test periods when 40 mM CaCl2, or MgCl2 (made isoosmolal with NaCl) was infused. Both CaCl2 and MgCl2 depressed water absorption in ileal segments by 30–40% compared with NaCl. In man, 6 ileostomy patients were given 60 mmole CaCl2 or MgCl2, orally over 24 h for 3 days in a randomized, cross-over trial. CaCl2 increased the average ileostomy output from 984 to 1,336 g/day; MgCl2 increased output from 939 to 1,352 g/day. A corollary of these results was that magnesium hydroxide would be a less effective laxative when this insoluble hydroxide is not converted to soluble chloride by gastric juice. To test this inference, 5 totally gastrectomized patients were given 60 mmole of Mg(OH)2 or MgCl2 daily. Stool weight increased by 136 g after MgCl2 but not after Mg(OH)2.