To evaluate the efficacy and safety of oral magnesium supplementation, with magnesium chloride (MgCl2), in the treatment of newly diagnosed depression in the elderly with type 2 diabetes and hypomagnesemia. Twenty-three elderly patients with type 2 diabetes and hypomagnesemia were enrolled and randomly allocated to receive either 50 mL of MgCl2 5% solution equivalent to 450 mg of elemental magnesium or Imipramine 50 mg daily during 12 weeks. Widowhood or divorce in the last six months, alcoholism, degenerative illnesses of the nervous central system, recent diagnosis of diabetes, previous or current treatment with antidepressants, chronic diarrhea, use of diuretics, and reduced renal function were exclusion criteria. Hypomagnesemia was defined by serum magnesium levels < 1.8 mg/dL and depression by Yasavage and Brink score > or = 11 points. The primary trial end point was the improvement of depression symptoms. At baseline, there were no differences by age (69 +/- 5.9 and 66.4 +/- 6.1 years, p = 0.39), duration of diabetes (11.8 +/- 7.9 and 8.6 +/- 5.7 years, p = 0.33), serum magnesium levels (1.3 +/- 0.04 and 1.4 +/- 0.04 mg/dL, p = 0.09), and Yasavage and Brink Score (17.9 +/- 3.9 and 16.1 +/- 4.5 point, p = 0.34) in the groups with MgCl2 and imipramine, respectively. At end of follow-up, there were no significant differences in the Yasavage and Brink score (11.4 +/- 3.8 and 10.9 +/- 4.3, p = 0.27) between the groups in study; whereas serum magnesium levels were significantly higher in the group with MgCl2 (2.1 +/- 0.08 mg/dL) than in the subjects with imipramine (1.5 +/- 0.07 mg/dL), p < 0.0005. In conclusion, MgCl2 is as effective in the treatment of depressed elderly type 2 diabetics with hypomagnesemia as imipramine 50 mg daily.