Plants produce a range of volatile organic compounds (VOCs), some of which are perceived by the human olfactory system, contributing for a myriad of flavors. Despite the importance of flavor for consumer preference, most plant breeding programs have neglected it, mainly due to the costs of phenotyping and the complexity of disentangling the role of VOCs on human perception. To develop molecular breeding tools aimed at improving fruit flavor, we carried out a target genotyping and VOC extraction of a blueberry population. Metabolite genome-wide association analysis (GWAS) was used to elucidate the genetic architecture, while predictive models were tested to prove that VOCs can be accurately predicted using genomic information. Historical sensory panel was considered to assess how the volatiles influenced consumer. By gathering genomics, metabolomics, and sensory panel, we demonstrated that VOCs: i) are controlled by a few major genomic regions, some of which harboring biosynthetic enzyme-coding genes; ii) can be accurately predicted using molecular markers; and iii) can enhance or decrease consumer's overall liking. Here we emphasized how the understanding of the genetic basis and the role of VOCs on consumer preference can assist breeders in developing more flavorful cultivars at a more inexpensively and accelerated pace.