Solute concentration and soluble dye studies inferring that preferential flow accelerates field-scale contaminant transport are common but flux measurements quantifying its impact are essentially nonexistent. A tile-drain facility was used to determine the influence of matrix and preferential flow processes on the flux of mobile tracers subjected to different irrigation regimes (4.4 and 0.89 mm h(-1)) in a silt loam soil. After tile outflow reached steady state either bromide (Br; 280 kg ha(-1)) or pentafluorobenzoic acid (PFBA; 121 kg ha(-1)) was applied through the irrigation system inside a shed (3.5 x 24 m). Bromide fluxes were monitored at an irrigation rate of 4.4 mm h(-1) while PFBA fluxes were monitored at an irrigation rate of 0.89 mm h(-1). At 4.4 mm h(-1) nearly one-third of the surface-applied Br was recovered in the tile line after only 124 mm of irrigation and was poorly fit by the one-dimensional convective-dispersive equation (CDE). On the other hand, the one-dimensional CDE fit the main PFBA breakthrough pattern almost perfectly, suggesting the PFBA transport was dominated by matrix flow. Furthermore, after 225 mm of water had been applied, less than 2% of the applied PFBA had been leached through the soil compared with more than 59% of the applied Br. This study demonstrates that the methodology of applying a narrow strip of chemical to a tile drain facility is appropriate for quantifying chemical fluxes at the small-field scale and also suggests that there may be a critical input flux whereby preferential flow is initiated.