The colinearity between wheat and rice chromosome segments facilitates the prediction of wheat gene order using rice as a template. Conservation between coding sequences in these two species has been used also to predict Triticeae ESTs gene structure. However, detailed sequence comparisons of orthologous wheat-rice regions have revealed extensive rearrangements, particularly in intergenic regions. Even closely related genomes, such as the A genomes of polyploidy and diploid wheat, show less than 25% sequence conservation in some intergenic regions. These fast changes are generally associated with insertions and deletions of repetitive elements. Changes in gene regions are not as rapid as in intergenic regions, but are still frequent enough to complicate the identification of Triticeae candidate genes using rice. Three out of the four positional cloning efforts completed in my laboratory revealed that the targeted Triticeae genes were absent in the colinear rice regions. Therefore, sequencing of the orthologous Triticeae regions is a necessary step in positional cloning projects to identify potential non-colinear genes. During the positional cloning of the earliness per se gene EPS-1 we identified an additional complication. The colinear region between rice and wheat EPS-1 region is part of an ancient and large duplication involving large segments of rice chromosomes 5 and 1. Duplicated rice genes were less similar to each other than to the closest wheat orthologue, suggesting that this duplication preceded the divergence between wheat and rice. This hypothesis was confirmed by the duplication in the corresponding wheat chromosomes one and three of 18 of these duplicated rice genes. Independent gene deletions after the divergence of the wheat and rice lineages explained the observed alternations of colinearity between rice chromosome 5 and wheat chromosomes 1 and 3. Since this ancient duplication involved most of the rice genome, it would likely affect significantly most comparative genomic studies.