OBJECTIVE Neonatal hypomagnesemia is defined as total magnesium (TMg) < or = 0.65 mmol/L (1.6 mg/dl). However, magnesium (Mg) deficiency and sufficiency overlap at serum values of 0.57 to 0.74 mmol/L (1.4 to 1.8 mg/dl). We hypothesized that (1) some infants with TMg < or = 0.65 mmol/L (1.6 mg/dl) have normal ionized Mg values (normal neonatal range 0.40 to 0.56 mmol/L (0.97 to 1.36 mg/dl)); (2) the dose (6.0 mg of elemental Mg/kg) used to correct hypomagnesemia does not lead to elevation of ionized Mg; (3) after intravenous magnesium sulfate infusion, ionized calcium increases in patients with low baseline ionized Mg and decreases in patients with normal baseline ionized Mg. METHODS We recruited 22 neonates with TMg < or = 1.6 mg/dl. They received intravenous sulfate (6 mg elemental Mg/kg) over a 1-hour period. Serum TMg, ionized Mg, and ionized Ca were measured before and after magnesium sulfate infusion. An ion-selective electrode was used to allow direct measurement of ionized Mg and ionized Ca. RESULTS Thirteen (59%) of 22 neonates with TMg < or = 0.65 mmol/L (1.6 mg/dl) had normal IMg. In 7 (31%) of 22 cases ionized Mg increased slightly above 0.56 mmol/L (1.36 mg/dl); the maximum value was 0.61 mmol/L (1.48 mg/dl). The change in ionized Ca concentrations and the baseline ionized Mg value were inversely correlated (r = -0.79; p < 0.0001). CONCLUSIONS (1) Measurement of ionized Mg should prevent overdiagnosis and treatment of hypomagnesemia. (2) The dose used in this study is safe. (3) Ionized Mg concentrations are inversely correlated to the response of ionized Ca concentrations to an Mg load.