Neutral endopeptidase (NEP) is a 94 kDa ectoenzyme of the proximal tubule brush border, physiologically released into the urine with apical membrane fragments. As proximal tubular atrophy was a histological hallmark of Chinese herbs nephropathy (CHN), this study firstly determined renal excretion of NEP in healthy control subjects (N = 31), in patients with CHN (N = 26) and in women having consumed Chinese herbs and whose renal function was normal but running the risk of developing CHN (N = 27). Another patient group consisted of female patients with glomerular diseases (N = 12). At the same time, measurements of urinary microproteins (Clara cell protein, retinol binding protein, beta 2-microglobulin and alpha 1-microglobulin) were performed, as indicators of tubular dysfunction. Cell damage was estimated by the excretion of N-acetyl-beta-D-glucosaminidase (NAG). In the control group, the physiological NEP enzymuria was 43.1 micrograms/24 hr (geometric mean). In CHN patients, levels of urinary NEP were significantly decreased in those with moderate renal failure (26.7 micrograms/24 hr; N = 21; P < 0.05) and almost abolished in end-stage renal failure patients (4.35 micrograms/24 hr; N = 5; P < 0.05). In patients at risk as well as in patients with glomerular diseases, urinary NEP levels were not statistically different from those observed in control subjects (40.68 micrograms/24 hr and 48.5 micrograms/24 hr, respectively). Several degrees of tubular dysfunction and injury were noted in patients groups, as attested by increased urinary microproteins and NAG excretions. Considering the data from control and CHN patients, NEP enzymuria positively correlated with individual creatinine clearance values (r = 0.76; P = 0.0001) and negatively correlated with urinary microproteins levels (r = -0.55; P = 0.00001). Finally, NEP was regularly quantitated in the urine of 6 CHN patients for a period ranging from six months to two years and in 19 patients at risk during two years, respectively. In the first group, renal function progressively deteriorated in 3 patients, leading them to renal replacement therapy after 38 to 115 weeks. Stable parameters were observed in the remaining 3 patients. A direct correlation between creatinine clearance and NEP excretion was found longitudinally in each case. In the second group, no significant change of urinary NEP levels was observed (45.9 micrograms/24 hr), in parallel with stable renal function. Taken together, these results indicate that, in CHN patients, NEP enzymuria provides a rapid and noninvasive determination of the degree of structural impairment affecting the proximal tubular population and further reflecting the severity of the renal disease. The interest of this urinary marker in monitoring the progression of other tubulointerstitial diseases remains to be assessed.