Abstract: In contrast to our understanding of genomic variation present within and between species, the quantitative characteristics of phenotypic diversity remain poorly defined. Here, we describe a comprehensive phenomics dataset of over 15,000 molecular and morphological traits collected in 22 genetically diverse strains of Saccharomyces cerevisiae. These phenotypes span a range of biomolecules and cellular characteristics, including digital gene expression levels, protein and metabolite abundances, and cellular morphology. Over 50% of all measured traits exhibited significant inter-strain variability, highlighting the pervasive influence of genetic variation on phenotypic differences between strains. We identified hundreds of robust associations between phenotypes and genetic variants, some of which persist across multiple levels of the phenotypic hierarchy. Finally, we show that phenomics is a powerful framework for predictive biology by integrating information across heterogeneous phenotypes to accurately infer protein abundances and morphological traits.