Recently, we discovered that the centre of the glomerulus is drained by a series of tributaries between podocytes forming an anastomosing system of channels or interpodocyte spaces (IPS) (Neal, 2003, 2005). The IPS could remove primary urine from the centre of the glomerulus to its periphery and thence into the shell like peripheral urinary space (PUS) which is similar to Bowmans original space of 1842 (Neal, 2005). We have now reconstructed the IPS and studied its dimensions more fully throughout the glomerulus to see how it changes under perfusion compared with non-perfusion conditions. Kidneys from humanely killed rats were either fixed with glutaraldehyde by immersion (n=3, renal arterial pressure of 0mmHg) or perfusion after a flush of the kidneys with Ringer solution (n=5, renal arterial pressure of 100mmHg throughout). Kidney pieces were postfixed, dehydrated, embedded and serial ultrathin sections of glomeruli were cut. The IPS were reconstructed from electron micrographs of these sections using the Reconstruct program (version 1.0.3.9, John C. Fiala). The general course of the IPS was followed by reconstructing the mean free path of the IPS in the glomerulus. This was complemented by measuring the widths of the IPS. Data were compared using unpaired t tests and Mann-Whitney U tests. The reconstructions showed that IPS branched by splitting into two and bridging podocytes often temporarily blocked the IPS. The width of the IPS in these selected reconstructions and also in randomly selected central and peripheral locations showed that the IPS was wider in perfusion fixed glomeruli (1.56±0.13μm, n=30, 1-2 glomeruli per rat) and narrower in immersion fixed (non-perfused) glomeruli (0.85±0.05μm, n=25, 1-2 glomeruli per rat; P=0.0001, t test or U test). While the perfusion pressure (100mmHg) and the absence of balancing colloid from the perfusate suggests that glomerular ultrafiltration should be artificially high, we expect that tubuloglomerular feedback will bring the glomerular filtration rate back to near normal prior to fixation. This suggests that the IPS inflates with fluid under near normal filtration conditions which may be controlled by the constrictions that exist where the IPS meets the PUS (Neal, 2005). The glomerulus inside the Bowmans capsule appears to work in a functionally distended fashion – much like the whole kidney.