Improving traits in wheat has historically been challenging due to its large and polyploid genome, limited genetic diversity and in-field phenotyping constraints. However, within recent years many of these barriers have been lowered. The availability of a chromosome-level assembly of the wheat genome now facilitates a step-change in wheat genetics and provides a common platform for resources, including variation data, gene expression data and genetic markers. The development of sequenced mutant populations and gene-editing techniques now enables the rapid assessment of gene function in wheat directly. The ability to alter gene function in a targeted manner will unmask the effects of homoeolog redundancy and allow the hidden potential of this polyploid genome to be discovered. New techniques to identify and exploit the genetic diversity within wheat wild relatives now enable wheat breeders to take advantage of these additional sources of variation to address challenges facing food production. Finally, advances in phenomics have unlocked rapid screening of populations for many traits of interest both in greenhouses and in the field. Looking forwards, integrating diverse data types, including genomic, epigenetic and phenomics data, will take advantage of big data approaches including machine learning to understand trait biology in wheat in unprecedented detail.