Personalized cancer medicine promises to determine why some individuals are at a greater risk for developing cancer and to also improve the outcome for individual cancer patients through the best choice of treatment. Ovarian cancer is a major cause of cancer deaths, yet there are few genetic risk factors beyond BRCA1 and BRCA2 identified. To identify cancer-associated genetic variation, most research to date has focused on using genome wide association studies (GWAS) or identifying DNA alterations within protein-coding portions of genes. However, cancer patients contain functional genetic changes in non-protein coding regions of cancer genes. For example, inherited DNA changes in microRNAs (miRNAs) and their binding sites are associated with an increased risk for developing cancer. MiRNAs are key regulators of gene expression that target many genes, including important cancer genes, by binding to 3’UTR control centers that regulate gene stability and use. Despite the critical role of miRNAs and their binding sites in cancer gene regulation, they have yet to be systematically examined for genetic variation. Here we captured DNA regions corresponding to thecomplete collection of validated human miRNAs and the 3’UTRs of about 6000 cancer-associated genes from 31 ovarian cancer patients and identified many known as well as many novel DNA alterations. Of note, multiple SNPs in the 3’UTR of the VEGF receptor/FLT-1, E2F2 and PCM1 oncogenes were found highly enriched in ovarian cancer patients compared to the 1000 Genome Project dataset. A Sequenom validation in a case-control study of ~400 individuals confirmed a novel variant in PCM1 is significantly associated with ovarian cancer (p value = 0.003). This work identifies a new ovarian cancer locus and suggests that non-coding regions of cancer patients should not be ignored in cancer resequencing efforts.