In this article, we highlight biomarkers for poly(ADP-ribose) polymerase inhibitor (PARPi) sensitivity and resistance and discuss their implications for the clinic. We review the predictive role of a range of DNA repair genes, genomic scars, mutational signatures, and functional assays available or in development. The biomarkers used for patient selection in the specific Food and Drug Administration-approved indications for breast, ovarian, prostate, and pancreatic cancer vary across tumor type and likely depend on disease-specific DNA repair deficiencies but also the specifics of the individual clinical trials that were conducted. Mutations in genes involved in homologous recombination and/or replication fork protection are synthetic lethal with PARPi. Cancers with homologous recombination deficiency exhibit high genomic instability, characterized by genome-wide loss of heterozygosity, among other genomic aberrations. Next-generation sequencing can identify multiple patterns of genomic changes including copy number variations, single-nucleotide variations, insertions/deletions, and structural variations rearrangements characteristic of homologous recombination deficiency. Clinical trial evidence supports the use of BRCA mutation testing for patient selection, and for ovarian cancer, there are 3 commercial assays available that additionally incorporate genomic instability for identifying subgroups of patients that derive different magnitudes of benefit from PARPi therapy. Finally, we summarize new strategies for extending the benefit of PARPi therapy toward broader populations of patients through the use of novel biomarkers. Ultimately, design of a composite biomarker test combining multiple mutational signatures or development of a dynamic assay for functional assessments of homologous recombination may help improve the test accuracy for future patient stratification.