Puromycin aminonucleoside (PAN) has been suggested to reduce glomerular charge density, to create large glomerular "leaks," or not to affect the glomerular barrier. Therefore, we analyzed glomerular charge and size selectivity in vivo and in isolated kidneys perfused at 8 degrees C (cIPK) in control and PAN-treated rats. The fractional clearances (theta) for albumin and Ficoll of similar hydrodynamic size were 0.0017 +/- 0.0004 and 0.15 +/- 0.02, respectively, in control cIPKs. Two-pore analysis gave similar results in vivo and in vitro, with small- and large-pore radii of 47-52 and 85-105 A, respectively, in controls. Puromycin increased the number of large pores 40-50 times, the total pore area over diffusion distance decreased by a factor of 25-30, and the small-pore radius increased by 33% (P < 0.001 for all comparisons of size selectivity and theta). The effect of PAN was less dramatic on the estimated wall charge density, which was 73% of that of controls. We conclude that puromycin effectively destroys the glomerular size barrier with minimal effects on charge density.