Genetically modified, herbicide-resistant crop (HRC) cultivars, which allow for simplified weed control decisions compared with conventional cultivars, have considerable potential in Latin America. The number of herbicide applications can be reduced in HRCs, and otherwise difficult-to-control species, including red rice in rice or herbicide-resistant weeds, can potentially be managed. The American tropics include the centers of origin of several crops, such as corn and potato, and natural or agrestal floras containing wild near relatives of introduced crops, including rice and cotton, for which HRCs could be used. Potential direct impacts of HRC adoption on biodiversity in Latin America include changes in the genetic diversity of crops, increased volunteer crop problems, and invasion by resistant cultivars of natural areas beyond the farm boundary. Additionally, there is a risk associated with the escape of transgenes from HRCs, involving introgression to weedy relatives on a field scale leading to amplification effects of existing weeds or modification of gene pools of crop progenitors in the centers of origin or diversity. Possible indirect effects of HRCs include the potential expansion of agriculture into uncleared wild areas made economically attractive by the more efficient cropping system, and adverse effects on nontarget organisms and ecosystem processes. The occurrence of spontaneous hybrids of crop–weed or near-relative complexes is known in the United States for rice and sunflower, and the development of HRC–weed hybrids would provide a particularly difficult set of weed management problems. For crops in which HRCs are currently in use or under development, the greatest risks in Latin America appear to be with corn, cotton, and potato. However, some of the genetic and geographical barriers reduce the risk of hybridization between these crops and their wild relatives. Furthermore, unlike the case of feral rapeseed on road verges in Europe, there is no known example of a conventionally bred crop or interfertile hybrid with a near relative becoming established outside of cultivation in Latin America. Soybean, the most widely adopted HRC in Latin America to date, is an exotic beyond the range of its wild relatives. With the exception of O. glumaepatula, there appears to be little threat to endemic wild species from HRC rice as strong infertility barriers should prevent transgene flow, but possible transgene movement to the conspecific red rice may become an issue. Beyond the field boundary, a key question concerns the possible persistence of HRC x wild relative hybrids and the manner in which resistance traits affect their overall fitness. Designing practical risk assessment protocols on which to base HRC release approval is a considerable challenge, and it is vital that Latin American states continue to build and maintain functional biosafety regulatory structures.