The fructosamine levels and some metabolic control parameters were evaluated in 120 diabetics and in 60 controls. Higher levels of fructosamine were found in diabetics compared to normals (3.22 +/- 0.06 vs. 1.82 +/- 0.05 mmol/l; P less than 0.001). Positive correlations were found between fructosamine and metabolic control parameters in diabetic subjects. Moreover, non-insulin-dependent diabetic subjects had better correlations. The study of fructosamine variation in relation to plasma glucose change showed that fructosamine levels were modified after at least 2 weeks of good metabolic control. In conclusion, the fructosamine assay could be a useful method for medium-term metabolic control evaluation.