We assessed plant interspaces in July 2007 using continuous line intercepts in twice-replicated pastures of northern mixed-grass prairie with contrasting grazing treatments: 1) long-term (25 yr) heavily grazed, dominated by the bunchgrass blue grama (Bouteloua gracilis), and 2) ungrazed, dominated by the rhizomatous grass western wheatgrass (Pascopyrum smithii). The number of plant interspaces was 26% higher in pastures heavily grazed, but the amount of soil surface occupied by plant interspaces was 27% greater without grazing. Plant interspaces were larger without grazing (14.8 ± 1.2 cm, mean ± 1 SE) than heavily grazed (8.9 ± 0.4 cm). Plant interspaces represented 87% and 68% of the total soil surface in the ungrazed and heavily grazed communities, respectively. The percentage of soil surface covered by plant interspaces < 20 cm was higher for the heavily grazed (94%) compared to the ungrazed (79%). Litter cover in the plant interspaces was higher without grazing (80 ± 1%) compared to the heavily grazed (57 ± 3%). Grazing-induced structural changes from a rhizomatous- to a bunchgrass-dominated vegetation community were manifest in the size and distribution of plant interspaces. Ecological consequences for erosion from raindrop impacts in larger plant interspaces in the ungrazed community are likely offset by greater litter cover in these communities; conversely, lower litter cover in heavily grazed pastures may increase erosion potential despite occurrence of smaller plant interspaces and less proportion of the soil surface covered by interspaces. Management practices that increase the cover of litter in plant interspaces should reduce the potential of erosion from water and wind in this semiarid rangeland.