BACKGROUND There is still controversy as to whether increased urinary albumin excretion (UAE) in patients with type 2 diabetes mellitus may have similar pathognomonic relevance as in type 1 diabetes and whether improved metabolic control may beneficially influence increased UAE in type 2 diabetic patients to the same extent as in type 1 diabetic patients. METHODS In a cross-sectional study in 234 patients with type 2 diabetes (age, 64 +/- 1 years, known duration of diabetes, 14 +/- 1 years) UAE, haemoglobin A1, blood pressure, cholesterol, triglyceride and creatinine levels were measured and signs of retinopathy were evaluated. Results were compared with the findings in 247 patients with type 1 diabetes (age, 39 +/- 1 years, duration of diabetes, 15 +/- 1 years). In a longitudinal study, UAE, haemoglobin A1, blood pressure, cholesterol, triglyceride and creatinine levels were measured in 41 patients with type 2 diabetes and secondary failure of oral hypoglycemic treatment (age, 62 +/- 1 years, known duration of diabetes, 11 +/- 1 years) before and after 2-year insulin treatment. RESULTS In the cross-sectional study, 39% of the type 2 diabetic patients had increased UAE, 27% had microalbuminuria, in contrast to 21% and 14%, respectively, of the type 1 diabetic patients (p < 0.01). In type 2 diabetes, macroalbuminuria was detected after an average of 15 +/- 2 years in contrast to 25 +/- 2 years in type 1 diabetes (p < 0.01). In comparison to macroalbuminuric type 1 diabetic patients, macroalbuminuric type 2 diabetes patients exhibited a lower prevalence of renal insufficiency (30 vs. 53%, p < 0.05) as well as of retinopathy (59 vs. 88%, p < 0.05), but a higher prevalence of hypertension (93 vs. 65%, p < 0.05) as well as of hyperlipidaemia (p < 0.01). There was a significant relation between UAE and haemoglobin A1 in patients with type 1 diabetes (p < 0.01) which could not be demonstrated in type 2 diabetic patients. In the longitudinal study, UAE fell from 86 +/- 28 to 51 +/- 16 mg/24 hours (p < 0.05) and the prevalence of UAE of more than 30 mg/24 hours decreased from 45 to 25% (p < 0.05) after 2-year insulin treatment, while blood pressure remained constant and haemoglobin A1 fell from 12.3 +/- 0.4 to 8.8 +/- 0.3% (p < 0.01). In comparison to patients with normal or normalized UAE after 2 years, patients with persistently elevated UAE had more advanced kidney disease (p < 0.05), more pronounced insulin resistance and a higher prevalence of calculated mean arterial blood pressure above 103 mm Hg and of haemoglobin A1 higher than 9% (normal, < 7%) in the course of the study (p < 0.05). CONCLUSIONS In patients with both type 2 and type 1 diabetes mellitus, increased UAE is a marker for the nephropathy-related sequelae of long-term hyperglycaemia. In type 2 diabetic patients, however, the frequently pre-existing hypertension as well as other disease mechanism of the "metabolic syndrome" ("syndrome X") may additionally determine course and progression of diabetic kidney disease.