The insulin-mimetic effect of vanadate is well established, and vanadate has been shown to improve insulin sensitivity in diabetic rats and humans. Although the exact mechanism(s) remain undefined, we have previously demonstrated a direct relation of intracellular free magnesium (Mg(i)) levels to glucose disposal, to insulinemic responses following glucose loading, and to insulin-induced ionic effects. To investigate whether the insulin-mimetic effects of vanadate could similarly be mediated by Mg(i), we utilized (31)P-nuclear magnetic resonance spectroscopy to measure Mg(i) in erythrocytes from normal (NL, n=10) and hypertensive (HTN, n=12) subjects, before and after incubation with insulin and with different doses of sodium vanadate. In NL, vanadate elevated Mg(i) levels, with maximum efficacy at 50 7 micromol/L (186+/-6 to 222+/-6 7micromol/L, P>0.01), as did physiologically maximal doses of insulin, 200 7microU/mL (185+/-6 to 222+/-8 7micromol/L, P<0.01). In HTN, only vanadate, but not insulin, increased Mg(i) (insulin: 173+/-7 to 180+/-9 7micromol/L, P=NS; vanadate: 170+/-7 to 208+/-10 7micromol/L, P<0.01). Mg(i) responses to insulin (r=0.637, P<0.001), but not to vanadate (r=0.15, P=NS), were closely and directly related to basal Mg(i) levels. We conclude that (1) both vanadate and insulin stimulate erythrocyte Mg(i) levels; (2) cellular Mg(i) responses to insulin, but not to vanadate, depend on basal Mg(i) content-the lower the basal Mg(i), the less the Mg(i) response to insulin. As such, (3) Mg(i) responses to vanadate were equivalent among HTN and NL, whereas HTN cells exhibited blunted Mg(i) responses to insulin, and (4) the ability of vanadate to improve insulin sensitivity clinically may be mediated, at least in part, by its ability to increase Mg(i) levels, which in turn, helps to determine cellular insulin action.