Abstract Purpose: Patient derived xenograft (PDX) models represent a platform for defining therapeutic opportunities based on their ability to maintain the genomic and molecular features of their respective tumors. We used a comprehensive genomics approach to analyze PDXs for oral squamous cell carcinoma (OSCC), which had not been previously described. This involved confirming that the PDXs conserve the heterogeneous somatic landscape of the primary tumors from which they were derived. Methods: We established a large panel of OSCC PDX models by engrafting fresh, primary tumor samples into NOD scid gamma (NSG) mice. We performed whole genome sequencing (WGS; n=10), whole exome sequencing (WES; n=15), and transcriptome sequencing (RNAseq; n=15) on matched primary and first passage xenografts. Somatic events, including mutations, copy number alterations (CNAs), loss-of-heterozygosity (LOH) regions, and structural variation (SV), were detected independently in tumors and PDXs. Conservation was defined by the correlation of variant allele frequency (VAF) distributions and gene expression patterns, presence of larger somatic events, and predictions made about the subclonal architecture of the tumor and xenograft. Results: PDXs were established with an overall success rate of 47% (56/118). The somatic landscape of primary tumors was representative of the published genomic data on OSCC, including amplifications of CCND1 and EGFR family members, inactivating mutations and loss of TP53 and CDKN2A, and chromosome 3q amplification. These somatic events were also strongly conserved in matched xenograft samples. Mutations that were detected in either the xenograft or primary tumor, but not its matched sample, consisted primarily of low-frequency mutations. The predicted clonal architecture, however, revealed correlative subclonal heterogeneity between matched xenograft and primary tumor. This includes maintenance of founding clone mutations and consistent VAF distributions, indicating that a heterogeneous tumor population engrafted. Conclusions: This comprehensive approach establishes early-passage PDXs as high fidelity models for OSCC. Their ability to maintain the genomic heterogeneity detected in primary OSCC tumors presents a clinically relevant platform for understanding tumor biology and identifying novel therapeutic strategies. Citation Format: Katie M. Campbell, Tianxiang Lin, Ashley E. Winkler, Paul Zolkind, Zachary L. Skidmore, Erica K. Barnell, Ian Hagemann, Elaine R. Mardis, Malachi Griffith, Rebecca D. Chernock, Obi L. Griffith, Ravindra Uppaluri. Oral cavity squamous cell carcinoma xenografts display conservation of primary tumor genomic heterogeneity [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3847. doi:10.1158/1538-7445.AM2017-3847