Abstract Premalignant tissue samples collected at different timepoints along the path to invasive tumors have been used to establish the typical order of somatic genetic events in several cancer types. For many other cancer types, however, precancer tissue is unobtainable, leaving the genetic progression unknown. Here, we demonstrate an approach that can reconstruct progression histories for such cancers from exome sequencing of primary tumors. We validated our recent computational method, PhylogicNDT, on classic, HPV-negative head and neck squamous cell carcinoma (HNSCC), recapitulating the previous experimentally determined sequence of genetic events that occur during progression. We uncovered progression of HPV-positive HNSCC tumors, which lack premalignant tissue, identifying subtype specific early drivers. We associated the timing of major changes in ploidy with the development of intra-tumor genetic heterogeneity and shorter overall survival. In cancers with unobtainable premalignant tissue, our approach can fill the knowledge gap about early progression, providing the basis for development of experimental models and methods for early detection, interception, and prognostication. Citation Format: Ignaty Leshchiner, Edmund A. Mroz, Justin Cha, Daniel Rosebrock, Oliver Spiro, Juliana Bonilla-Velez, William Faquin, Armida Lefranc-Torres, Derrick Lin, William Michaud, Gad Getz, James W. Rocco. Establishing the early genetic progression of cancers with unobtainable premalignant disease [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2712.