Weed seed predation by carabid beetles (Coleoptera, Carabidae) may contribute to weed control in crop fields but to date, evidence of weed regulation by carabids is scarce. Predation rates are driven by several factors (crop management techniques, habitat quality, pedoclimate, carabids’ abundance and activity and their weed seed preferences). Models are needed to evaluate the impact of weed seed predation on multi-annual weed dynamics, and thus on crop production and weed biodiversity. We developed a predation model which was added to an existing weed dynamics model (FlorSys, Colbach et al., 2014). The model was built from literature data and previous experiments. It simulates the effects on daily predation rates of management techniques, vegetation cover, climate, carabids’ intra-annual abundance variation and in-field weed seed preferences of carabids. A sensitivity analysis of predation to model parameters was run to identify the parameters that have the most impact on predation and are thus required to be accurately estimated. Simulations were run to evaluate the impact of weed seed predation on crop production and biodiversity in ten cropping systems. The cropping systems were based on a rapeseed/wheat/barley rotation, differing in rotational crop diversity, herbicides, ploughing and tillage frequency. Including predation by carabid beetles in the simulations only marginally changed multiannual weed dynamics. Effects of predation on crop production and weed biodiversity were rare and depended on the cropping system. Predation decreased flora species richness in the ten cropping systems because the model assumed that carabids predated their preferred weed species rather than the most frequent ones. To correct this bias, we plan to include the effect of seed density dependence on predation to the model. We will simulate a wider range of cropping systems to conclude on the contribution of seed predation by carabids in the biological regulation of weeds