The effects of the biological cross-linker transglutaminase, the neutrophil oxidant hydrogen peroxide, and neutrophil proteinases on glomerular basement membrane permeability have been examined using an in vitro model of glomerular ultrafiltration. The main focus of the study lies in determining whether any of the test agents were able to render glomerular basement membrane more permeable to protein. Guinea pig liver transglutaminase was used as a model enzyme to test for the effect of biological cross-linkers on glomerular basement membrane permeability. It cross-linked glomerular basement membrane proteins, caused membrane contraction, and rendered glomerular basement membrane less permeable both to water and the low molecular weight protein marker myoglobin but had no effect on the membrane permeability to the high molecular weight marker protein bovine serum albumin or serum protein. The pathophysiological relevance of the effect is discussed. Hydrogen peroxide increased glomerular basement membrane permeability to water and proteins but the effect depended on hydrogen peroxide concentration and incubation time. The minimum concentration needed to render glomerular basement membrane more permeable to bovine serum albumin and serum protein was 1 M and the minimum incubation time needed was 6 hrs. A respiratory burst analysis of activated neutrophils showed that the average concentration of hydrogen peroxide that could be generated by the neutrophils was less than 50 mM and the time taken for extracellular hydrogen peroxide concentration to fall off to zero was less than 1 hr. Therefore, neutrophils seemed unable to generate and sustain a sufficiently high hydrogen peroxide concentration to render glomerular basement membrane more permeable to protein in vivo. Proteinases extracted from pig neutrophil granules were used to assess their effect on glomerular basement membrane permeability. The extract showed activity against glomerular basement membrane and the activity was primarily attributed to the serine proteinases elastase and cathepsin G, judged from substrate and inhibitor analyses. The proteinase extract also contain latent metalloproteinases, activatable by the organomercurial 4-aminophenyl mercuric acetate and calcium ions. Once activated, they also showed activity against glomerular basement membrane. The extract rendered glomerular basement membrane more permeable to water, myoglobin, bovine serum albumin, and serum protein. The increase in membrane permeability to water and proteins was due to membrane thinning and an increase in the intrinsic porosity of the membrane. When the serine and metalloproteinases were allowed to act in concert, they synergistically degraded glomerular basement membrane and increased the membrane permeability to serum protein and water. The study provides the first direct evidence that pathophysiological amounts of serine and metalloproteinases are able to render glomerular basement membrane more permeable to protein and suggests they may be capable of promoting proteinuria in neutrophil-dependent forms of immune glomerulonephritis.