Solute transportwas studied in a variable-charge soil (Nitisol)with two following pig manure applications. There were three identical soil columns (diameter=37.5 cm; soil depth=85 cm) equippedwith TDR probes and tensiometers, one of which served as an untreated control. Dispersivities inferred using the CXTFIT 2.1 code presented inverse patterns with depth for nitrate and chloride breakthrough curves, while the normalized intensities had the same patterns with depth for both applications. For nitrates, the retardation factors steadily decreased with depth from 4.85 and 3.57 at 17 cm depth to 2.1 and 1.86 PV (pore volume) at 85 cm depth for each column, respectively. For chlorides, the retardation factor increased lineary with depth, from 1.05 at 17 cm depth to 1.76 and 1.86 PV at 85 cm depth. After the first application, the mean difference between nitrate and chloride retardation factorswas 3.8 PV at 17 cm depth and it regularly decreased to 0.34 PV at 85 cm depth. For the second application, the mean difference was 2.52 PV at 17 cmdepth and it regularly decreased to 0 PV at 85 cm depth. The kinetics of nitrate production by the pig manure nitrification process modified the pH, the ionic strength of the soil solution and then the anionic exchange capacity. This could explain the high nitrate retardation factors until 55 cm depth and the difference between nitrate and chloride retardation factors. The earlier chloride adsorption at anionic exchange sites, combined with a selectivity coefficient to the detriment of nitrates, counterbalanced the delay in nitrate production due to the kinetic mineralisation of pig manure. Nitrate fluxes then caught up with the chloride fluxes at the outlet.