The overarching goal of exercise genomics is to illuminate exercise biology and behaviour in order to better understand the preventive and therapeutic values of exercise. An ancillary aim is to understand the role of genomic variation in human physical attributes and sports performance. The aim of this report is to briefly comment on the current status of exercise genetics and genomics and to suggest potential improvements to the research agenda and translational activities. First, the genomic features of interest to the biology of exercise are defined. Then, the limit of the current focus on common variants and their implications for exercise genomics is highlighted. The need for a major paradigm shift in exercise genomics research is discussed with an emphasis on study designs and appropriately powered studies as well as on more mechanistic and functional research. Finally, a summary of current practices in translational activities compared with what best practice demands is introduced. One suggestion is that the research portfolio of exercise genomics be composed of a larger fraction of experimental and mechanistic investigations and a smaller fraction of observational studies. It is also recommended that research should shift to unbiased exploration of the genome using all the power of genomics, epigenomics and transcriptomics in combination with large observational but preferably experimental study designs, including Mendelian randomisation. In all cases, emphasis on replications is of paramount importance. This represents an extraordinary challenge that can only be met with large-scale collaborative and multicentre research programmes.