We have developed a capture antibody, noncompetitive, enzyme-linked immunoassay for urinary apolipoprotein A-I (Apo A-I) in urine, with use of affinity-purified polyclonal antisera against Apo A-I. A 96-well microtiter plate format is used, with unconcentrated urine as sample and dilutions of serum or high-density lipoprotein (HDL) as standards. The intra- and interassay variation (CV) averaged 7.4% and 9.4%, respectively. The limit of detection is low (1.25 ng/L), and no cross-reactivity with Apo B, C, E, or A-II was detected. The mean (+/- SD) concentrations of Apo A-I in urine of patients with glomerular proteinuria were a thousandfold greater (38.4 +/- 23.1 mg/L) than in normal subjects (16.3 +/- 11.3 micrograms/L in men, 17.97 +/- 7.7 micrograms/L in women, a significant difference, P less than 0.001). Apo A-I measurements correlated very well (r = 0.92) with selectivity index assessment. The diurnal variation of the concentration of Apo A-I in urine appears to result from dilution related to fluid intake. This enzymatic method is easy to perform, can be used with large numbers of samples, and is adaptable for use in the routine clinical laboratory. The method holds promise for discriminating between normal and subclinical kidney disease populations by measuring the concentrations of urinary Apo A-I excreted on HDL particles.