Eragrostis plana (Poaceae) is a perennial grass introduced from South Africa to the state of Rio Grande do Sul in southern Brazil. Currently, it is considered an invasive grass in several regions of the world, including South America, where it has caused negative ecological and socio-economic impacts. Ecological niche models, using bioclimatic variables, are often used to predict the potential distribution of invasive species. In this study we prepared two bioclimatic models for E. plana using the Genetic Algorithm for Rule-set Production, the first based on data from its native region (South Africa) and the second on data from both the native and invaded (South America) regions. We then projected each model onto South America to identify regions vulnerable to invasion by the species, and compared our results with available records of the species in South America. Finally, we explored the model's predictions for the existence of a bioclimatic niche shift during the invasion process of E. plana in South America, using multivariate statistical analysis. The model created with native distribution data was only able to predict (with highly suitable habitat) the region of introduction of E. plana in South America. However, the current distribution, as well as the region of introduction of the species, was reliably predicted by the model created with data from both native and invaded regions. Our multivariate analysis supports a hypothesis of bioclimatic niche shift during the invasion process of E. plana in South America.