Because their is a total body deficit of magnesium in malnutrition, magenesium salts are routinely prescribed in treatment. Several children were observed to develop severe hyperchloraemic acidosis after admission to the ward, and urine from recovering children was found to be highly acidic. This study tested the hypothesis that the acidosis was iatrogenically induced by the magnesium chloride supplement. Six malnourished and 5 recovering children were treated in the standard way except that they were given magnesium chloride (1 mmol/kg/d) for 4 days and then magnesiumacetate for 4 days. The order of the salts was changed in alternate children. Jugular venous Astrup and timed urine collections were made at the end of each period. The malnourished children developed sytemic acidosis on the magnesium chloride supplement (pH 7.3 ñ 0.01 vs - #.6 ñ 1.1). Those childrenwho were given the acetatesalt first remained in acid-base equilibrium whereas those who were given the chloride first had not regained normal acid base status after 4 days on the acetate. On the chloridesalt the urinary pH was lower (5.0 ñ 0.3 vs 6.5 ñ 0.3), the titratable acidity higher (1.5 ñ 0.2 vs 0.4 ñ 0.3 æEq/min) and the total acid (TA = ammonia - HCO3) output higher (5.5 ñ 0.8 vs 2.1 ñ 0.4 æEq/min), than on the acetatesalt. The recovering children also developed a mild systemic acidosis on the chloride (Base excess - 5.5 ñ 0.9 vs 2.3 ñ 0.5) and excreted a highly acidurine, pH 4.8 ñ 0.1; on the acetate the urine was also acidpH 5.1 ñ 0.1. The total acidity output was 20.4 ñ 1.1 æEq/min on the chloride and 15.3 ñ 1.4æEq/min on the acetate. It is concluded that magnesium chloride is the cause of the iatrogenic systemic acidosis seen in our malnourished children. Whilst malnourished they excrete a maximally acidurine, however the total acid output is small and leads to a progressively positive hydrogen ion balance, which the child cannot correct within four days of ceasing the chloride load. This work was supported by the Wellcome Trust (AU)