Abstract Purpose: Multiple melanoma GWAS have identified significant associations between single nucleotide polymorphisms (SNPs) and melanoma. As multiple variants within a given genetic locus may contribute to disease risk in different individuals, a novel computational approach to analyzing GWAS data by ‘wide-locus’ region has been shown to increase analytical power, and, in other diseases, has led to the discovery of new susceptibility loci. To date, this approach has not been applied to the study of germline genetic risk factors for melanoma. We applied this method to publically available melanoma GWAS data to identify novel germline risk factors for melanoma. Methods: We re-analyzed data from the M.D. Anderson Cancer Center melanoma GWAS study (dbGap accession: phs000187.v1.p1). In the original study, 1,977 cases and 1,047 controls of European ancestry were genotyped on an Illumina OMNI1-Quad chip. We age-matched cases and controls using “age-censoring” to account for differences in melanoma risk across different age groups. Biostatistical analysis of ‘wide-loci’ was performed across all SNPs to evaluate risk loci based on the influence of neighboring SNPs. The re-analysis was performed by individuals blinded to specific results of the original study. Genome-wide significance for each wide locus was based on number of regions examined, and was set at p<10−6.16. Results: Using the ‘wide-locus’ approach, we identified 21 risk loci on 16 chromosomes that met genome-wide significance. Firstly, we replicated all of the genome-wide significant findings of the original, single-SNP analysis, identifying specific SNPs at HERC2 and wide loci at MC1R and AFG3L1 that met genome-wide significance. Secondly, we identified additional SNPs not replicated to genome-wide significance in the original analysis but previously discovered in other populations. These additional SNPs reside in the MC1R, OCA2, CPNE7, DBNDD1, and AFG3L1 loci. Finally, we identified 15 new candidate loci meeting genome-wide significance, with the most significant findings located at KLK12 (associated with carcinogenesis in other cancers), PDE8 (involved with cAMP intra-cellular signaling and immune activation), and LY86 (involved in immune cell activation). Conclusion: Applying ‘wide-locus’-based statistical analysis to melanoma GWAS data replicated the findings of prior single-SNP analyses and identified multiple, new candidate risk loci worthy of further investigation. Citation Format: Rachel M. Cymerman, Benedetta Bigio, Martin P. Seybold, David Polsky, Knut M. Wittkowski. A novel computational re-analysis of published GWAS data suggests new risk loci for melanoma susceptibility. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4627. doi:10.1158/1538-7445.AM2015-4627