In eight aged non-insulin-dependent diabetes mellitus (NIDDM) subjects, insulin response and action were studied before and after chronic magnesium supplementation (2 g/day) to diet. Chronic magnesium supplementation to diet versus placebo produced 7) a significant increase in plasma (0.83 ± 0.05 vs. 0.78 ± 0.06 mM, P < .05) and erythrocyte (2.03 ± 0.06 vs. 1.88 ± 0.09 mM, P < .01) magnesium levels, 2) an increase in acute insulin response (AIR) (4.0 ± 0.6 vs. - 1 .6 ± 0.6 mU/L, P < .05) to glucose pulse, and 3) an increase in glucose infusion rate (GIR) (3.6 ± 0.6 vs. 2.9 ± 0.5 mg kg∼1 min−1P < .025) calculated in the last 60 min of a euglycemic-hyperinsulinemic (100U m2 · min−1 during 180 min) glucose clamp. Net increase in AIR, glucose disappearance rate after glucose pulse, and GIR were significantly and positively correlated to the net increase in erythrocyte magnesium content calculated after chronic magnesium supplementation to diet. In conclusion, our data suggest that NIDDM subjects may benefit from therapeutic chronic administration of magnesium salts.