The aim of this study was to investigate the relationship between serum, erythrocyte and urine magnesium levels and retained magnesium percentage in newly diagnosed diabetic children. In a cross-sectional study, 34 children with insulin dependent diabetes mellitus (IDDM) and 21 healthy age- and sex-matched control subjects were screened for their serum, erythrocyte, and urine magnesium levels. Magnesium tolerance test was performed on diabetic and control subjects. Serum and erythrocyte magnesium levels in diabetic children were significantly lower than in healthy controls (plasma magnesium, p<0.05; erythrocyte magnesium, p<0.001); however, serum magnesium level was in normal range in diabetics and controls. Erythrocyte magnesium levels in diabetic children showed an inverse correlation with percentage of retained magnesium load (r=-0.44, p<0.01). Urine magnesium excretion in diabetic children (7.12 +/- 2.18 mmol/g creatinine/24-hr) was significantly higher than in healthy controls (4.0 +/- 1.35 mmol/g creatinine/24-hr) (p<0.001). There was a negative correlation between erythrocyte magnesium (2.07 +/- 0.62 mmol/L) and urine magnesium (7.12 +/- 2.18 mmol/g creatinine/24-hr) (r=-0.68 p<0.01) in diabetic children. Magnesium tolerance test showed that percentage of retained magnesium in diabetic children (66 +/- 26%) was significantly higher than in controls (16 +/- 7%) (p<0.001). This study is the first study to simultaneously investigate serum, erythrocyte and urine magnesium levels and magnesium tolerance test in newly diagnosed diabetic children. In conclusion, erythrocyte magnesium levels decrease earlier than serum magnesium in diabetic children. The follow-up parameters in diabetics may include the policy of monitoring magnesium status. Erythrocyte magnesium measurement is preferred to serum magnesium. Magnesium tolerance test is a reliable and sensitive method, which may be used as an alternative to erythrocyte magnesium measurement or in combination with it in hospitalized diabetic children.