Seventy-six normal, free-living subjects were given supplements of 200 micrograms chromium (Cr) in the form of chromic chloride or a placebo in a double-blind crossover study with 3-month experimental periods. Twenty of the 76 subjects had serum glucose concentrations greater than or equal to 100 mg/dL 90 minutes after a glucose challenge (1 g glucose per kilogram of body weight). Chromium supplementation significantly decreased (P less than 0.05) the 90-minute glucose concentration of these subjects from 135 +/- 9 to 116 +/- 11 mg/dL; fasting glucose concentrations also decreased significantly. The 90-minute serum glucose levels of the 35 subjects with glucose concentrations less than the fasting serum glucose level were increased significantly by Cr supplementation, from 71 +/- 1 to 81 +/- 4 mg/dL. Fasting and 90-minute serum glucose concentrations of the remaining subjects who displayed 90-minute glucose concentrations greater than fasting levels but less than 100 mg/dL were not affected by Cr supplementation. In this study, immunoreactive serum insulin concentration, body weight, lipids, and other selected clinical variables did not change significantly during Cr supplementation. These data demonstrate that Cr supplementation decreases the serum glucose levels of subjects with 90-minute glucose concentrations greater than or equal to 100 mg/dL following a glucose challenge, increases serum glucose levels of subjects with 90-minute glucose concentrations less than fasting levels, and has no effect on the serum glucose levels of subjects with 90-minute glucose values similar to but greater than fasting levels.