Albuminuria is a dominant biochemical feature of developing diabetic nephropathy. A disturbed metabolism of heparan sulphate characterized by an increased loss of anionic charges in the basement membrane has been considered as one of the main factors causing an increased albumin output into urine. All therapeutic approaches inducing a reduction of the albumin excretion rate (AER) have a protective effect on renal function. The effect of glycosaminoglycan sulodexide on albuminuria was studied in a group of 53 diabetic patients (26 Type 1 and 27 Type 2) with micro and macroalbuminuria. Sulodexide (Vessel Due F) was administered intramuscularly in one daily dose (600 lipasemic units) for 3 weeks followed by a 6 week wash-out period. A significant decrease of AER was found in a total cohort of patients following just 1 week of sulodexide treatment (mean 162 micrograms/min, range 10-2708 micrograms/min vs mean 248 micrograms/min, range 20-3160 micrograms/min, P < 0.001). This effect lasted 3-6 weeks after drug withdrawal. Similar results were obtained if Type 1 and Type 2 diabetic patients were evaluated separately but a delay of the AER reduction was observed in the latter group. In all patients the mean AER was reduced to 60-65% of the initial values. A greater effect of sulodexide on albuminuria was observed in patients with AER above 200 micrograms/min than in those with microalbuminuria (a reduction to 47 vs 65% of the initial output). Sulodexide did not significantly reduce albuminuria in 28% of diabetic patients ('non-responders'). In conclusion, glycosaminoglycan sulodexide may reduce AER in patients with micro or macroalbuminuria and it could slow down development of diabetic nephropathy.