Objective: Build an agent-based model of Spartina plant growth on a simulated mud flat environment in order to study the effects of spatial pattern and initial seedling placement on the invasion dynamics of the population. Background: Invasive non-indigenous plant species can have potentially serious detrimental effects on local ecosystems and, as a result, costly control efforts often have to be put in place to protect habitats. An example of an invasive problem on a global scale involves the salt marsh grass species from the genus Spartina. The spread of S. anglica in Europe and Asia has drawn much concern due to its ability to convert coastal habitats into cord-grass monocultures and to alter the native food webs. Materials & Methods: We present here a new model framework in C++ that can be used to explicitly represent complex spatial and temporal patterns of plant invasions in order to be able to predict quantitatively the impact of these factors on invasion dynamics. This would be a useful tool for assessing eradication strategies and choosing optimal control solutions in order to be able to minimise future control costs. Results: Simulations were carried out to explore different hypothetical eradication/control strategies for Spartina infestations on an open mudflat environment. The growth of patches is strongly influenced by the shape of the infestation and this modelling approach can be used to quantitatively predict the impact of these factors on invasion dynamics. © 2012 EUROSIS-ETI.