The Ramipril Efficacy in Nephropathy Core and Follow-Up Study found that > or =36 mo of continued ramipril therapy decreased substantially the risk of end-stage renal failure (ESRF) in patients with chronic nephropathies and a urinary protein excretion rate > or =3 g/24 h. This study investigates the time-dependent changes in GFR in these patients and in control subjects who were randomized to conventional therapy during the Core period and switched to ramipril during the Follow-Up study. Analyses included 150 patients (continued ramipril: n = 74; switched to ramipril: n = 76) who had at least three GFR measurements (including baseline) during the whole observation period and a subgroup of 43 patients (continued ramipril: n = 26; switched to ramipril: n = 17) who had at least six GFR measurements, including at least three on the Core and at least three on the Follow-Up study. Ramipril (1.25 to 5 mg/d) and conventional therapy were targeted at achieving a diastolic BP below 90 mm Hg. The main efficacy variables were GFR and ESRF (need for dialysis). Analysis was by intention to treat. Throughout the study, the mean +/- SEM rate of GFR decline (deltaGFR) was significantly lower in patients continued on ramipril compared to those switched to ramipril (0.51+/-0.09 versus 0.76+/-0.10 ml/min per 1.73 m2 per mo, P<0.03). In patients on continued ramipril who had at least six GFR measured--but not in control subjects--deltaGFR progressively improved with time and, in the cohort with the longest follow-up, decreased from (in ml/min per 1.73 m2 per mo): 0.16+/-0.12 (at 18 mo) to 0.10+/-0.05 (at 60 mo). This rate was about 10-fold slower compared to patients on conventional therapy during the REIN Core study. Analyses of the individual slopes found that at the end of the follow-up, 10 of 26 patients on continued ramipril therapy had a positive deltaGFR and another 10 patients had an improvement of deltaGFR while on ramipril therapy. DeltaGFR significantly improved in parallel with a significant reduction in proteinuria. Changes in deltaGFR (P = 0.0001) and proteinuria (P = 0.04) were significantly different in the two groups. Baseline characteristics and changes in systolic and diastolic BP and 24-h urine urea and sodium excretion were comparable. The present results offer evidence that in chronic nephropathies, the tendency of GFR to decline with time can be effectively halted, even in patients with remarkably severe disease.