To determine the specificity of the urine excretion of albumin as a measure of glomerular permeability in early insulin-dependent diabetic nephropathy, the effect of variable glomerular filtration and urine flow rates on albumin, beta 2-microglobulin excretion, and the fractional renal clearance of neutral dextran (Stokes Einstein Radius 24-46 A) was examined. Five insulin-dependent diabetic subjects with normal glomerular permeability (albumin excretion less than 30 micrograms/min) and one with elevated albumin excretion (195 micrograms/min) were studied pre and post strict glucose control with constant subcutaneous insulin infusion for 7 days. The albumin excretion in the 5 subjects never exceeded 30 micrograms/min during wide variations in glomerular filtration and urine flow rates. A positive correlation between beta 2-microglobulin excretion and urine flow (r = 0.81), and glomerular filtration (r = 0.77) rates was observed. In contrast, albumin excretion showed no correlation, indicating different factors affect the excretion rate of albumin and beta 2-microglobulin. Therefore, elevated albumin excretion (greater than 30 micrograms/min) in insulin-dependent diabetes is due to increased glomerular permeability and not changes in glomerular filtration and urine flow rates, and the albumin/ beta 2-microglobulin ratio may not be a valid indicator of changing glomerular permeability. The fractional neutral dextran clearances remained unchanged with variation in glomerular filtration and urine flow rates. The sieving curve was identical in all subjects for neutral dextran 40 A, the size of albumin, suggesting that reduced glomerular charge selectivity may contribute to increased albuminuria in progressive diabetic glomerulosclerosis.