For the past four decades, genetic investigations have largely focused on the discovery of high impact genetic variants that drive disease pathogenesis. This endeavour has been remarkably successful, but a critical knowledge gap remains: why do individuals who carry the same disease-predisposing genetic variant often display strikingly different phenotypes? The mechanisms underlying this common phenomenon, known as variable penetrance, are poorly understood. This limits our understanding of disease biology and has important implications for genetic counselling and therapeutics. In this thesis, I use large-scale genomic and functional genomic datasets to explore the determinants of variable penetrance in the context of genetic ophthalmic disorders. First, I interrogate the relationship between gene expression variability and variable penetrance and demonstrate that genes enriched in variants with inconsistent phenotypic consequences are often more variable in terms of gene expression. I then further explore this concept in the context of albinism, an archetypal recessive disorder associated with reduced melanin pigment in the eyes and skin. Using haplotype-based analysis, I find that a common non-coding variant modulates the penetrance of a protein-coding, albinism-implicated change. I also show that the joint analysis of common and rare coding and non- coding variation helps to identify a molecular diagnosis in a significant proportion of previously unexplained albinism cases. Finally, I investigate the molecular mechanisms underpinning the marked phenotypic variability of a developmental eye disorder, North Carolina macular dystrophy. I show that disease-implicated non-coding variants impact distinct enhancer regions that have tissue-specific activation patterns correlated with the associated clinical phenotypes. Combined, these results advance our understanding of variable penetrance and have important implications for diagnostic genetic testing and genetic counselling.