One of the concerns raised over the introduction of genetically modified crops is that transgenes will invade populations of wild relatives, causing ecologically significant changes in fitness. In recent years, this has given rise to several studies estimating hybridization rates and the fitness of crop-wild relative hybrids. These studies have established that transgenes are likely to move to F1 hybrids, albeit at low frequency. Hybridization, however, is not synonymous with introgression, and questions remain as to whether particular transgenes will cause ecologically significant changes in recipient plant populations. Research effort should now focus on estimating any changes in the fitness of a population as a consequence of having a transgene, understanding genotype x environment interactions, and deducing the extent to which pathogens and herbivores (transgene targets) regulate wild relative populations. This will involve a combination of manipulative experiments and empirically motivated mathematical models.