Aims: Lymph node metastasis (LNM) is one of the strongest predictors of worse outcome in oral squamous cell carcinoma (OSCC) [PMID: 24064974]. However, it remains unclear whether nodal dissemination is encoded within the primary tumor or emerges later through interactions with the microenvironment and lymphatic niche. While most studies focus on coding genes, we asked whether integrated analysis of non-coding genomic alterations, chromatin accessibility, and RNA expression could identify primary tumor programs that predispose OSCC to LNM. Methodology: We performed whole genome, transcriptome, and ATAC sequencing on 60 primary OSCC tumors, 34 with and 26 without LNM [PMID: 40185094]. LNM-associated non-coding alterations were linked to nearby chromatin and transcriptional changes. Unsupervised and supervised modeling defined an LNM-associated multi-omic signature, followed by external validation in TCGA-OSCC, derived from the TCGA head and neck cancer cohort. Results: We identified a tuned genome-chromatin-RNA signature that distinguished LNM-positive from LNM-negative primary tumors in the discovery cohort (AUC=0.894; p=2.17 × 10⁻⁷), driven mainly by genome and RNA features with supportive chromatin input. Bootstrap stability analysis refined the model to a robust 7-gene RNA readout linked to vesicle/adaptor signaling, Rho-GTPase regulation, kinase signaling, and ubiquitin-mediated transcriptional control. The refined RNA+CNV alteration score validated in TCGA-OSCC for discrimination of pathologic node status (AUC=0.671; p=2.87 × 10⁻⁴). Conclusion: These findings suggest that LNM potential is partly encoded within primary OSCC through non-coding alterations. Multi-omic integration identified a reproducible LNM-associated signature, highlighting candidate regulatory loci and RNA markers that may improve biological insight and clinical risk stratification. Ongoing work is expanding the cohort to 200+ samples and validating these findings in independent datasets.