Compound-specific nitrogen isotope analysis was shown to be a promising tool for the quantitative assessment of abiotic reduction of nitroaromatic contaminants (NACs) under anoxic conditions. To assess the magnitude and variability of 15N fractionation for reactions with dissolved reductants, we investigated the reduction of a series of NACs with a model quinone (anthrahydroquinone-2,6-disulfonate monophenolate; AHQDS-) and a Fe(II)-catechol complex (1:2 Fe(II)-tiron complex; Fe(II)L2(6-)) over the pH range from 3 to 12 and variable reductant concentrations. Apparent kinetic isotope effects, AKIEN, for the reduction of four mononitroaromatic compounds by AHQDS- ranged from 1.039 +/- 0.003 to 1.045 +/- 0.002 (average +/- 1sigma), consistent with previous results for various mineral-bound reductants. 15N fractionation for reduction of 1,2-dinitrobenzene and 2,4,6-trinitrotoluene by AHQDS- and that of 4-chloronitrobenzene by Fe(II)L2(6-), however, showed substantial variability in AKIEN-values which decreased from 1.043 to 1.010 with increasing pH. We hypothesize that the isotope-sensitive and rate-limiting step of the overall NAC reduction can shift from the dehydration of substituted N,N-dihydroxyanilines (large 15N fractionation upon N-O bond cleavage) to protonation or reduction of nitroaromatic radical anions (small 15N isotope effect upon electron transfer) consistent with calculations of semiclassical 15N isotope effects. Our results imply that a quantitative assessment of NAC reduction using compound-specific isotope analysis (CSIA) might need to account for homogeneous and heterogeneous reactions separately.