Is there any risk that the threshold for admixture of genetically modified seeds in the harvest of a conventional cultivar, 0.9% in Europe, will be exceeded in the case of inbreeder crops? Using herbicide-resistant foxtail millet, Setaria italica, as a model of a preferentially autogamous crop, such as wheat and rice, field experiments show that genotype admixture due to pollen flow between adjacent fields is about 0.03% on average for the 10 adjacent meters, and 10 times less in the next 20-m lane. In the case of a maternally inherited resistance gene, the admixture rate is at least 100 times lower. Recessive herbicide resistance has also been tested but would be efficient only if the agreed coexistence rules were based on phenotype detection.