Understanding how genetic variants drive phenotypic differences is a major challenge in molecular biology. Single nucleotide polymorphisms form the vast majority of genetic variation and play critical roles in complex, polygenic phenotypes, yet their functional impact is poorly understood from traditional gene-level analyses. In-depth knowledge on the impact of single nucleotide polymorphisms has broad applications in health and disease, population genomic or evolution studies. The wealth of genomic data and available functional genetic tools make Drosophila melanogaster an ideal model species for studies at single nucleotide resolution. However, to leverage these resources and potentially combine it with the power of functional genetics to enhance genotype-phenotype research, it is essential to develop techniques to predict functional impact and causality of single nucleotide variants. Here, we present FlyCADD, a functional impact prediction tool for single nucleotide variants in D. melanogaster. FlyCADD, based on the Combined Annotation-Dependent Depletion (CADD) framework, integrates over 650 genomic features - including conservation scores, GC content, and DNA secondary structure - into a single metric reflecting a variants predicted impact on evolutionary fitness. FlyCADD provides impact prediction scores for any single nucleotide variant on the D. melanogaster genome. We demonstrate FlyCADDs utility with some examples of application, including the ranking of phenotype-associated variants to prioritize variants for follow-up studies, evaluation of naturally occurring polymorphisms, and refining of CRISPR-Cas9 experimental design. FlyCADD provides a powerful framework for interpreting impact of any single nucleotide variant in D. melanogaster, thereby improving our understanding of genotype-phenotype connections.