Denitrification is sensitive to changes in soil physical properties that affect solute transport, air content and gas diffusion. Using lysimeters, containing intact soil from intensively tilled (IT) and no-tilled (NT) soil used to grow forage crops, we examined how simulated animal treading at different moisture contents (above and below field capacity; >FC and <FC respectively) affected losses of nitrous oxide (N2O), dinitrogen (N-2) and nitrate (NO3). We applied N-15-labelled NO3 (250kg N ha(-1)) to the soil surface after treading (applied at 220 kPa to 40% of the soil surface), or to untrodden soil. Drainage occurred following weekly application of water over the experiment (two pore volumes over 84 days). Treading at >FC greatly increased denitrification, especially from IT soil and produced the greatest amount of N-2 (64kg N ha(-1)), N2O (8.2kg N ha(-1)), as well as the lowest N2O to N2O+N-2 ratio (0.08) and NO3 leaching (136kg N ha(-1) below 30cm). In both the uncompacted or compacted soils <FC, emissions of N2O were greater (1.5-2.7% of N applied) and the N2O to N2O+N-2 ratios were closer to 0.2 compared to compaction at >FC. Treading at <FC had minimal or no effect on denitrification compared to untrodden soil. Fluxes of N-2 and N2O were strongly influenced by the weekly irrigation-drainage cycle. The N-2 production and reduction in NO3 leaching were best correlated with increases in microporosity and reduced saturated hydraulic conductivity following treading. Although recovery of N-15 was high (84.3%), the remainder of the balance was likely lost as either N-2 or, of greater concern, as N2O. Practically, animal trampling on wet soils, especially when recently cultivated, should be avoided.