Past research has shown that changes in grazing-resistance traits may be associated with genetic changes in plant populations. Little is known about spatial genetic relationships within plant populations (spatial genetic structure) and any grazing effects on these relationships. Here we present observations of the fine-scale spatial genetic structure in three grass species in semiarid environments (Arizona, Mexico, and Argentina). In each environment, populations of a dominant grass species were sampled from two sites with contrasting livestock grazing histories. Plant genotypes were described with the use of amplified fragment length polymorphism markers. In Arizona, populations of sideoats grama (Bouteloua curtipendula var. caespitosa Gould and Kapadia) differed in that one has never experienced livestock grazing, whereas cattle have grazed the other. In the other two environments, populations exposed to long-term heavy grazing were examined, along with those that experienced either moderate grazing (Mexico, blue grama [Bouteloua gracilis {Willd. ex Kunt} Lag. ex Griffiths]) or extended exclusion of livestock (Argentina [Poa ligularis Nees ex Steud.]). Based on independent analysis of each population, we observed no differences in average gene diversity between populations of each species. With the use of analysis of molecular variance we found slight but significant genetic differentiation between populations with different grazing histories in Arizona and Argentina. Significant genetic structure was present in all populations and indicated an inverse relationship between spatial and genetic distance. Interestingly, this relationship was most pronounced in the cattle-free sideoats grama population, suggesting larger genetic neighborhood areas in the absence of livestock. Less distinct differences in spatial genetic structure associated with grazing history were evident in the other two species. We hypothesize that livestock grazing may lead to increased homogeneity in genetic structure at the landscape scale. Effectively examining this hypothesis presents many experimental challenges.