Depletion of intracellular free magnesium (Mg-i) is a characteristic feature of insulin resistance in essential hypertension, but it is not cleat to what extent low Mg-i levels contribute to insulin resistance, result from it, or both. As insulin-like growth factor I(IGF-I) may improve insulin resistance, we investigated whether this peptide could similarly improve Mg-i responsiveness to insulin in hypertension, and whether this effect was related to any direct IGF-I effect on Mg-i. P-31-Nuclear magnetic resonance spectroscopy was used to measure Mg-i in erythrocytes from 13 fasting normotensive and 10 essential hypertensive subjects before and 30, 60, and 120 min after incubation with a physiologically maximal dose of insulin (200 mu U/mL) and with different doses of recombinant human IGF-I (0.1-100 nmol/L). In normotensive subjects, IGF-I elevated Mg-i (P < 0.05) in a dose- and time-dependent fashion, as did insulin (P < 0.05). However, in hypertensive subjects, maximal Mg-i responses to insulin, but not to IGF-I, were blunted [insulin, 163 +/- 11 to 177 +/- 10 mu mol/L (P = NS) IGF-I, 164 +/- 6 to 190 +/- 11.7 mu mol/L (P < 0.05)]. Furthermore, for insulin, but not for IGF-I, cellular Mg-i responsiveness was closely and directly related to basal Mg-i levels (insulin: r = 0.72; P < 0.01; IGF-I: r = 0.18; P = NS). Lastly, blunted Mg-i responses to insulin could be reversed by preincubation of hypertensive cells with EGF-I. We conclude that 1) both IGF-I and insulin stimulate erythrocyte Mg-i levels; 2) cellular Mg-i responses to insulin, but not to IGF-I, depend on basal Mg-i levels, i.e, the higher the Mg-i the greater the sensitivity to insulin; and 3) IGF-I potentiates insulin-induced stimulation of Mg-i at doses that themselves do not raise Mg-i. These effects of IGF-I may underlie at least in part its ability to improve insulin sensitivity clinically. Together, these data support a role for IGF-I in cellular magnesium metabolism and emphasize the importance of magnesium as a determinant of insulin action.