Gene flow and introgression from crops to wild species can modify the adaptive potential and weediness of arable-land plant species. It gained recent importance with the release of genetically modified (GM) crops because of the risk of transfer of herbicide-resistance genes to related weeds. It is also a permanent possibility that could have occurred in the past with any conventional variety. In order to investigate this “normal” phenomenon and its impact, we developed three approaches in the framework of the study of gene flow between oilseed rape (Brassica napus) and wild radish (Raphanus raphanistrum). First, we compared the growth and reproduction of progeny of advanced backcross generations of interspecific hybrids (stabilized BC9 at 2n=18 chromosomes) displaying various levels of molecular markers of the crop parent. Lower fitness accompanies the presence of introgression. Second, we compared morphology and growth of weedy populations of wild radish belonging to regions where the crop was present or absent for long time. Evidence for neutral introgression in wild radish is presented. Third, we developed a simulation model to predict the fate of an advantageous transgene in a population of wild radish following hybridization with a GM crop. We present the generic model structure, comprising stage-structure (from F1 to BC2), spatial realism (i.e. including field borders and waste places) and stochasticity. By focusing on the most influencing parameters, we show that the large uncertainty on input parameters led to unpredictability for the fate of the transgene in the wild radish population.