To evaluate the impact of systemic blood pressure and albuminuria on the progression of diabetic nephropathy, we followed 41 IDDM patients with persistent albuminuria (> 300 mg/24 h) by measuring glomerular filtration rate (51Cr-EDTA technique), blood pressure, and albuminuria. None of the patients were taking drugs other than insulin. Arterial blood pressure, albuminuria, and blood glucose were measured four to eight times/yr, whereas glomerular filtration rate was determined twice yearly. During the median investigation period of 36 (15-66) mo, glomerular filtration rate decreased from 102 +/- 23 to 83 +/- 27 ml.min-1 x 1.73 m-2 (P < 0.001), albuminuria increased from 633 to 1435 micrograms/min (P < 0.001), and blood pressure rose from 133/85 +/- 10/9 to 149/93 +/- 8/11 mmHg (P < 0.001). Univariate analysis revealed a significant correlation between the rates of decline in glomerular filtration rate and diastolic blood pressure (r = 0.52, P < 0.01) and glomerular filtration rate and albuminuria (r = 0.34, P < 0.02). But stepwise multiple linear regression analysis only showed a significant correlation between the rate of decline in glomerular filtration rate and diastolic blood pressure (P < 0.01). In patients with diastolic blood pressure below the mean value of 89 mmHg, stepwise multiple regression analysis showed that albuminuria and not blood pressure was correlated significantly with rate of decline in glomerular filtration rate. Patients were stratified by average value of diastolic blood pressure measured during the investigation period.(ABSTRACT TRUNCATED AT 250 WORDS)