Grazing of livestock on pasture land can result in changes to the physical condition of soil, particularly as a result of trampling and changed organic matter status of the topsoil. Over time, changes to pasture botanical composition may also occur, which might further affect soil structure. The extent of the effects, and the rate of soil and pasture recovery when livestock are removed, will depend on the grazing management tactics employed. A field study was established at Orange, New South Wales, to compare soil physical properties under 'set-stocked' grazing of Merino sheep, 'high intensity--short duration' rotational grazing, an ungrazed control, and pasture cages. Topsoil bulk density, hydraulic conductivity, and organic carbon content were measured annually over 3 years, and image analysis of soil macroporosity was carried out annually to quantify changes in pore geometry. Only the topsoil macropore properties changed significantly between treatments over the 3 years. In particular, the structural quality of the topsoil under set-stocked grazing changed, as indicated by a decrease in total macroporosity and a smaller proportion of macropores. In contrast, stable structural conditions were maintained under rotational grazing. Possibly the best soil structure for plant growth, represented by high values of total macroporosity and macropore surface area, and a large range of pore sizes, was exhibited under the pasture cages, where pasture defoliation occurred in the absence of hoof pressure. It is concluded that grazing tactics are an important factor in the dynamics of soil macroporosity and the vertical continuity of macropores, as a result of the combined effects of hoof pressure and root channel development.