Until now, the renal clearance index of IgG to IgG4 (IgG/IgG4) as well as pancreatic to salivary amylase (PA/SA) were separately used as parameters of renal charge selectivity in diabetic and non-diabetic albuminuria. The suitability of the IgG index may be questioned because urinary loss of IgG rather reflects a size selective defect. In contrast, the amylase index seems more appropriate to reflect renal charge selectivity because its molecular size is comparable to albumin. We questioned whether IgG/IgG4 and PA/SA reflect renal charge selectivity in a comparable way in subjects with non-diabetic albuminuria over a wide range. Renal fractional clearances of albumin, IgG, IgG4, PA and SA were estimated from ambulatory 24-h urine samples in 12 subjects with normo-albuminuria (UAE 4 3-17 micrograms/min), six with micro-albuminuria (UAE: 14736-200 micrograms/min), and 20 with macro-albuminuria (UAE: 2301 608-13611 micrograms/min). Macro-albuminuria is associated with a reduced IgG/IgG4 and PA/SA, whereas micro-albuminuria is only associated with a reduced IgG/IgG4 compared to normo-albuminuria. A reduction of IgG/IgG4 (r = -0.75, P < 0.001) and PA/SA (r = -0.52, P < 0.001) correlates with an increased albuminuria. In addition, IgG/IgG4 correlates with PA/SA in the total population (r = 0.49, P < 0.01). IgG/IgG4 (r = 0.51, P < 0.05) correlates with the size selective index IgG/albumin in an opposite way to PA/SA (r = -0.52, P < 0.05) in 20 subjects with macro-albuminuria. Multiple regression analysis revealed IgG clearance to be the variable which contributes to the variance of albuminuria clearance for the greater part in our population. Both charge selective indices do not appear to correlate in micro-albuminuria. In addition, the presence of a size selective defect has a opposing effect on both charge selective indices. Although the reduction of IgG/IgG4 and PA/SA with increasing albuminuria suggests a progressive charge selective defect, albuminuria in our population is almost entirely explained by urinary loss of IgG. These data seriously question whether either one or both charge selective indices IgG/IgG4 and PA/SA do specifically reflect glomerular charge selectivity.