Studies of renal perfusion when kidney function tests are still normal could be useful to understand the pathophysiology of functional kidney impairment in cirrhosis; currently, this requires invasive methodology. Duplex Doppler ultrasonography allows noninvasive evaluation of intrarenal arterial resistances. In 19 nonascitic and 35 ascitic cirrhotic patients with normal kidney function (normal serum creatinine and urea levels) and in 17 controls, we measured the intrarenal arterial pulsatility index (Pulsatility index = [Peak systolic velocity -Minimum diastolic velocity]/Mean velocity) and the resistive index (Resistive index = [Peak systolic velocity -Minimum diastolic velocity]/Peak systolic velocity) by duplex Doppler ultrasonography after visualization of interlobar, interlobular or arcuate arteries by color Doppler ultrasonography. The pulsatility index and resistive index (calculated as the mean of three to five consecutive determinations) were significantly higher in cirrhotic patients than in control patients (pulsatility index: 1.16 f 0.24 vs. 0.78 +-0.05 [mean 2 S.D.], p < 0.001; resistive index: 0.67 f 0.06 vs. 0.53 2 0.03, p < 0.001). The pulsatility index and resistive index were significantly higher in nonascitic cirrhotic patients than in control patients, in ascitic patients than in nonascitic patients, in ascitic patients not treated with diuretics than in nonascitic ones and in ascitic patients treated with diuretics than in those not treated. The pulsatility index and resistive index measured before and after treatment with diuretics (7 to 15 days) in seven patients were significantly increased by treatment. The pulsatility index and resistive index were significantly higher in Child-Turcotte-Pugh class B and C patients than in class A patients. The pulsatility index and resistive index inversely correlated with creatinine clearance and directly correlated with heart rate. Thus renal arterial resistance indexes evaluated by duplex Doppler ultrasonography were