Recently, we were able to modify the glomerular charge barrier using perfusates with low and normal ionic strengths keeping the osmolality unchanged. The concentration of fixed charges was reversibly reduced from 35 to 12 mEq L-1 as the solution with low content of NaCl was introduced with no apparent effect on the size selectivity. It can be argued however, that the mannitol used for maintenance of osmolality may induce changes in glomerular permeability per se. To explore this possibility, isolated kidneys were perfused at 8 degrees with hyperosmolal mannitol solutions (560 mOsm) and compared with those perfused with standard albumin solutions (295 mOsm). The vascular resistance (PRU100) fell from 0. 14 +/- 0.01 to 0.11 +/- 0.01 mmHg min 100 g mL-1 as the mannitol solution was introduced (P < 0.001). As the blood pressure should remain unchanged, the flow was increased from 8 to 11 mL min-1. The glomerular filtration rate (GFR) increased by 50% from 320 +/- 40 to 490 +/- 20 microL min-1 g-1 (P < 0.001). Despite these changes in haemodynamical parameters, there was no significant change in the fractional clearance for albumin. Kidneys perfused with the mannitol solution showed well-preserved histology, while there was a conspicuous collapse of the cortical tissue and signs of tubular epithelial swelling with the standard perfusate. Moreover, all glomeruli were perfused in the mannitol group, as revealed by fluorescence of FITC dextran, while the distribution was uneven in the control kidneys. We conclude that perfusion of isolated kidneys with a hyperosmolal mannitol solution increased GFR by increasing the number of functionally active nephrons with no apparent effect on the glomerular barrier, a pattern differing from alteration of ionic strength.