Since Donald proposed ideotype breeding in 1968, agronomists and breeders have endeavoured to develop ideotype-oriented breeding methods to improve the dryland wheat production. Several scholars have discussed methods to reduce growth redundancy and improve the ideotype of crops, and have made some progress on dryland wheat breeding strategies and model innovation for cultivation management. Nevertheless, the evaluation of crop ideotype is not the same in different ecological zones, and progress on a specific crop ideotype for a particular ecological zone is not universally accepted. In this paper, the recent progress in ideotype research for the dryland wheat is reviewed and the evolutionary trend in growth redundancy in roots, stems and tiller number studied. At the same time, based on the data that has been collected under rainfed agricultural systems on Loess Plateau over the past two decades, we conducted regression analysis between root biomass, above-ground biomass, height and yield. On the Loess Plateau, spring wheat still has a high root/shoot ratio and removal of some root biomass can improve seed production, implying that there is a high incidence of growth redundancy in modern crop cultivars. The excessive growth of individual organs results in decreased yields. Further ecological mechanism of growth redundancy and interaction between ideotype and growth redundancy were also discussed. Research has shown that growth redundancy can be gradually eliminated in ideotype succession process, but can't be completely eliminated. A certain extent of the growth redundancy is the basic material for the ideotype. Over the long course of evolution from wild diploid and tetraploid, to modern hexaploid genotypes, spring wheat has been evolving from the high root/shoot ratio to the low root/shoot ratio, and this crop type could contribute to the increased yield of hexaploid wheat. We point out that the direction of dryland wheat ideotype breeding in the future should focus on the interaction between genotype of the wheat and its environment, reducing the intensity of competition among individuals and size inequality, promoting photosynthesis products that are deposited in the grain, and finally improving the population production. In summary, the interaction between a dryland wheat ideotype and growth redundancy is very complex. Selecting a dryland wheat ideotype should be based on the growth redundancy theory. Future breeding should be based on the basic principles of ecology, coupling the genotype and phenotype of dryland wheat, and then trade-off selection. The research focuses should move from research dealing with drought-resistance physiological genetics to the current ones dealing with crops ideotype. We believe that the theoretical and applied perspectives for the improvement of wheat production in the future will encompass the following: (1) The combination of the various morphological-physiological traits into one plant type; (2) The increase in the above-ground biomass and hence harvest index by selecting for small root systems and optimum resource allocation patterns; and (3) Selection and maintenance of weak competitors in population. Finally, the cooperative strategy that maximizes the population's grain production, although clearly desirable in agriculture, is not evolutionarily stable.