10523 Background: Melanoma has recently been suggested to be a highly heritable cancer with high twin-twin concordance. In contrast, prior studies on patients with familial atypical multiple mole melanoma (FAMMM) syndrome, multiple primary melanomas and early age of onset, all known to be associated with germline CDKN2A/CDK4 pathogenic variants, suggest that individuals with melanoma who carry germline alterations are rare. We studied the overall prevalence of germline cancer predisposition in a large prospective oncologic cohort with comparison to other defined cohorts of individuals with melanoma with available germline data. Methods: Individuals who presented to medical oncology clinic with a diagnosis of melanoma and personal or family history of multiple cancers were offered germline testing with a commercially available next generation sequencing panel (Invitae, San Francisco, CA). Eligibility criteria required ≥ 2 melanomas in an individual or family; melanoma and other cancer(s) in an individual; melanoma and at least 2 other cancers in 1^st- or 2^nd-degree relatives; age ≤35 at diagnosis; or limited family structure. Comparative analysis of germline NGS data from 3 additional selected and non-selected melanoma datasets was performed. Results: In a cohort of 400 oncology patients with melanoma who consented to commercial germline testing of 85 cancer-associated genes, a germline pathogenic/likely pathogenic (gP/LP) positive rate of 15.3% (n = 61) was observed. Genes previously associated with inherited melanoma ( BAP1, BRCA1/2, CDKN2A, MITF, TP53) comprised less than one-third of gP/LP variants (20, 32.7%); the majority of germline variants were in cancer predisposition genes not traditionally associated with melanoma (e.g. BRIP1, CHEK2, MSH2, PMS2, MLH1, RAD51C, BLM). Family history of non-cutaneous cancer (42, 69%) and personal history of melanoma with ≥ 1 non-cutaneous cancer (22, 36%) were the most common eligibility criteria met in gP/LP variant carriers. Analysis of germline data from other large oncologic and dermatologic melanoma datasets yielded gP/LP variant positive rates of 10.6% in an unselected oncologic melanoma cohort (TCGA, n = 470), 15.8% in a selected commercial testing cohort (Invitae, n = 12,571), and 14.5% in a highly selected, primarily dermatologic subset (Boston-Athens, n = 289). Conclusions: In oncologic and dermatologic cohorts, germline testing of selected individuals with melanoma yields rates of clinically impactful P/LP variant detection which exceed consensus standards for pretest probability. Most P/LP variants were found in genes associated with non-cutaneous cancers. Obtaining a family and personal history of cancer, particularly non-cutaneous cancers, and referring for broad panel-based germline testing in all individuals with melanoma are recommended.