Previous studies of the early autologous phase of bilateral nephrotoxic serum nephritis (NSN) in rats showed that whole kidney and single nephron glomerular filtration rate (GFR) were maintained at normal levels despite a 60% reduction in the product of surface area and hydraulic permeability (Kf). Factors responsible for this compensation were an increase of net ultrafiltration pressure, due primarily to an increased glomerular capillary pressure (PGC). This study was designed to investigate some possible causes of the compensation. Rats with bilateral NSN and normal GFR had an increased extracellular fluid volume (ECFV) 2 weeks after induction of NSN; control subjects did not change. To determine whether this ECFV expansion was responsible for triggering the compensation, we developed a unilateral NSN model with one diseased and one normal kidney. Unilaterally diseased rats did not experience an increase of ECFV. Values of Kf were 0.069 +/- 0.012 nl sec-1 mm Hg-1 in control subjects, 0.037 +/- 0.005 in bilateral NSN, and 0.043 +/- 0.006 in unilateral NSN. The elevation in PGC was the same in unilateral NSN as in bilateral NSN subjects and the same was true for the hydrostatic pressure difference across glomerular capillaries (delta P). Furthermore, in paired measurements on both kidneys of rats with unilateral NSN, PGC was significantly higher in the unilaterally diseased kidney than in the nondiseased kidney; sham control subjects had no difference. These results are interpreted to indicate that the signal that causes elevation of net ultrafiltration pressure is not a consequence of a systemic effect of NSN, but arises within the diseased kidney itself. To determine whether that signal involved some change in the mechanisms mediating autoregulation measurements were made of the response of whole kidney GFR and RBF to acute changes in arterial BP. Control rats and rats with NSN autoregulated both GFR and blood flow equally well. Tubuloglomerular feedback was studied by microperfusing loops of Henle and measuring proximal stop-flow pressure and early proximal flow rate. Stop-flow pressure was 4.0 mm Hg higher in rats with NSN at a loop perfusion rate of 10 nl/min, approximately the same difference that was found by direct measurement of PGC, but the sensitivity of response to changes in perfusion rate was the same in NSN as in control subjects. Finally, end proximal tubule flow rate was higher in NSN than in control subjects reflecting decreased proximal reabsorption. Thus, a normal feedback mechanism receives a signal that should cause afferent arteriolar constriction in NSN rats.(ABSTRACT TRUNCATED AT 400 WORDS)