Existing cancer-associated variant databases contain biases arising from duplicate entries and the inclusion of targeted sequencing panels, which interfere with accurate estimation somatic mutation frequency in cancer cohorts. To address this, we developed the Database of Recurrent Mutations (DORM), a web resource derived exclusively from whole-genome and whole-exome sequencing data. By filtering out targeted screens and non-recurrent variants, our analysis reveals that mutation recurrence significantly correlates with oncogenic activity, loss of tumor suppressor function, and unfavorable patient prognosis. In a pan-cancer analysis of EGFR, DORM identified frequent mutations outside the kinase domain that are underrepresented in other databases. This resource offers a streamlined, unbiased platform for mutation frequency analysis, enhancing biomarker discovery and the assessment of clinical variant significance.