Serum fructosamine levels were investigated in patients with uremia undergoing various modes of treatment. The serum fructosamine levels correlated positively with the blood glucose levels determined a week or two earlier. The fructosamine levels were significantly affected by the protein concentration, and those corrected for protein concentrations had a closer correlation to the blood glucose levels than did the uncorrected levels. The corrected fructosamine levels were not significantly different between healthy volunteers and nondiabetic patients with uremia on conservative treatment. In an in vitro system, fructosamine concentrations were hardly affected by urea, which is known to influence the level of hemoglobin A1. These results suggest that serum fructosamine measurement can provide us with reliable information on a short-term glycemic condition, even in azotemic patients. To be more precise, the serum level of fructosamine corrected for protein concentration can be an excellent glycemic index which is not susceptible to over- or dehydration and is of high clinical value, especially in the management of diabetic patients with chronic renal failure.