Vascular Endothelial Growth Factor (VEGF) potently increases systemic microvascular permeability to water (Bates & Curry, 1996), and is produced by glomerular cells (podocytes) (Bailey et al. 1999). We have refined the methods of Savin and colleagues (Savin & Terreros, 1981) to examine the effect of VEGF on the filtration characteristics of whole isolated glomeruli ex vivo. Adult male Wistar rats were anaesthetised with 5% halothane, and killed by cervical dislocation. Glomeruli were isolated using a standard sieving technique with mammalian Ringer solution containing 1% bovine serum albumin (BSA), and incubated in either control or VEGF (1nM)-containing solution for up to 3 hours. Glomeruli were individually loaded onto a suction micropipette and exposed to a flowing superperfusate of 1%BSA (in Ringer solution) at 37°C. Switching the superperfusate to a solution containing a higher BSA concentration (4-8%; in Ringer solution; Fig. 1) created a transglomerular oncotic gradient. Consequent fluid efflux caused a reduction in glomerular volume, which was recorded on videotape and measured off-line. The initial rate of change of glomerular volume was used to calculate glomerular ultrafiltration coefficient (LpA). A significant correlation between initial glomerular volume (V_i) and L_pA was noted (p<0.005; Spearman r = 0.45; n=47); henceforth L_pA values are corrected for V_i (L_pA/V_i; min^-1.mmHg^-1). Control L_pA/V_i values failed to display normal distribution. A linear relationship was demonstrated between the initial rate of volume change per unit V_i [(Δvol/Δt)/V_i] and the magnitude of the oncotic gradient applied (p<0.001; Spearman r = 0.59; n=28) (figure 1). L_pA/V_i values of glomeruli exposed to VEGF for 56 (mean) ±5 (S.E.M.) minutes were significantly higher [1.95±1.58; median±interquartile range (IQR); n=9] than those exposed to control solution (0.94±0.89; n=10) (p<0.01, Mann Whitney). Neither VEGF exposure for 15 minutes {paired L_pA/V_i: baseline 1.04 (0.77-1.96) [median (range)] vs VEGF 0.87 (0.60-2.03); p>0.6, Wilcoxon; n=5 pairs} nor 30 seconds [paired L_pA/V_i: baseline 1.07±0.49 (median±IQR) vs VEGF 0.98±0.43; p>0.7, Wilcoxon; n=10 pairs] elicited a rise in L_pA/V_i over baseline. These results show that prolonged exposure to exogenous VEGF can increase the ultrafiltration coefficient of renal glomeruli, ex vivo.