Background. Administration of prednisolone causes an abrupt rise in proteinuria in patients with a nephrotic syndrome. Methods. To clarify the mechanisms responsible for this increase in proteinuria we have performed a placebo controlled study in 26 patients with a nephrotic syndrome. Systemic and renal haemodynamics and urinary protein excretion were measured after prednisolone and after placebo. Results. After i.v. administration of 125-150mg prednisolone total proteinuria increased from 6.66 +/- 4.42 to 9.37 +/- 6.07 mg/min (P < 0.001). By analysing the excretion of proteins with different charge and weight (albumin, transferrin, IgG, IgG(4) and beta(2)-microglobulin) it became apparent that the increase of proteinuria was the result of a change in size selectivity rather than a change in glomerular charge selectivity or tubular protein reabsorption. Glomerular filtration rate rose from 83 +/- 34 ml to 95 +/- 43 ml/min (P < 0.001) after 5 h, whereas effective renal plasma flow and endogenous creatinine clearance remained unchanged. As a result filtration fraction was increased, compatible with an increased glomerular pressure, which probably contributes to the size selectivity changes. Since cortico-steroids affect both the renin-angiotensin system and renal prostaglandins, we have evaluated the effects of prednisolone on proteinuria after pretreatment with 3 months of the angiotensin-converting enzyme inhibitor lisinopril or after 2 weeks of the prostaglandin synthesis inhibitor indomethacin. Neither drug had any effect on prednisolone-induced increases of proteinuria. Conclusions. Prednisolone increases proteinuria by changing the size selective barrier of the glomerular capillary. Neither the renin-angiotensin axis nor prostaglandins seem to be involved in these effects of prednisolone on proteinuria.