The importance of measuring microalbuminuria is well established; however, controversy still exists regarding the type of urine specimen to be used for detecting early renal impairment of diabetic patients. To evaluate practical aspects, albumin concentration and albumin/creatinine ratio of first void urine samples as well as urinary albumin excretion in timed specimens were determined by immunoturbidimetric method 3 times within 3 weeks in 192 adult diabetic patients (136 men, 56 women; Type 1/Type 2 diabetes: 90/102; age: 51.4+/-10.8 yr; duration of diabetes: 15.3+/-9.1 yr; body mass index: 27.9+/-4.6 kg/m2; HbA1c: 8.54+/-1.46%; actual blood pressure: 138+/-14/82+/-8 mmHg; serum creatinine: 94+/-20 pmol/l; x+/-SD). According to the urinary albumin excretion values one-third of patients (31.2%-30.7%-34.4%) were normoalbuminuric (<30 mg/24 hr), more than half of the patients (55.8%-57.3%-53.6%) proved to be microalbuminuric (30-300 mg/24 hr), while the remaining group of patients (13.0%-12.0%-12.0%) was macroalbuminuric (>300 mg/24 hr). Comparing the results of successive measurements, good correlation was found between the same laboratory values (urinary albumin excretion: kappa=0.64, kappa=0.67; urinary albumin concentration: kappa=0.60, kappa=0.62; albumin/creatinine ratio: kappa=0.54, kappa=0.61; first vs second and second vs third measurements, respectively). The percentage of patients being in the same range of albuminuria (ie normo-, micro- or macroalbuminuria) at successive measurements was 79.7-81.2% with urinary albumin excretion values, 77.1-77.6% with urinary albumin concentration and 74.5-78.6% with albumin/creatinine ratio. Good correlation was found between urinary albumin excretion and urinary albumin concentration (kappa=0.54; 0.54; 0.57) and nearly the same correlation was observed between urinary albumin excretion and albumin/creatinine ratio (kappa=0.49; 0.47; 0.54) at the three consecutive measurements (n=192). Using values of urinary albumin excretion for comparison at all measurements, 79.3% sensitivity and 69.5% specificity were found for urinary albumin concentration, whereas 74.6% sensitivity and 68.8% specificity were documented for albumin/creatinine ratio. Beside the standard measurement of urinary albumin excretion in timed urine samples, the use of the more convenient morning urinary spot collection could also provide useful results (urinary albumin concentration or albumin/creatinine ratio) for detecting early renal involvement in diabetic patients.