One source of potential harm from the cultivation of transgenic crops is their dispersal, persistence and spread in non-agricultural land. Ecological damage may result from such spread if the abundance of valued species is reduced. The ability of a plant to spread in non-agricultural habitats is called its invasiveness potential. The risks posed by the invasiveness potential of transgenic crops are assessed by comparing in agronomic field trials the phenotypes of the crops with the phenotypes of genetically similar non-transgenic crops known to have low invasiveness potential. If the transgenic and non-transgenic crops are similar in traits believed to control invasiveness potential, it may be concluded that the transgenic crop has low invasiveness potential and poses negligible ecological risk via persistence and spread in non-agricultural habitats. If the phenotype of the transgenic crop is outside the range of the non-transgenic comparators for the traits controlling invasiveness potential, or if the comparative approach is regarded as inadequate for reasons of risk perception or risk communication, experiments that simulate the dispersal of the crop into non-agricultural habitats may be necessary. We describe such an experiment for several commercial insect-resistant transgenic maize events in conditions similar to those found in maize-growing regions of Mexico. As expected from comparative risk assessments, the transgenic maize was found to behave similarly to non-transgenic maize and to be non-invasive. The value of this experiment in assessing and communicating the negligible ecological risk posed by the low invasiveness potential of insect-resistant transgenic maize in Mexico is discussed.