Coccidioidomycosis (CM), caused by Coccidioides spp., accounts for a significant proportion of community-acquired pneumonia in endemic regions such as the southwestern United States. However, CM remains underdiagnosed, in part because of the limitations of existing diagnostic assays. There are likely uncharacterized antigens with diagnostic potential in the Coccidioides spp. proteome, however identifying these antigenic proteins from the large and poorly-understood proteomic space poses a significant challenge. PepSeq is a highly-multiplexed serology platform that measures antibody reactivity against a broad range of fully-customizable DNA-barcoded peptide antigens. Here, we develop a 244,000-plex PepSeq library providing proteome-wide representation of Coccidioides spp. We utilize this library to characterize protein and epitope-level antibody reactivity in CM patients, naturally infected pigtail macaques, and healthy human and animal controls. Our analysis revealed a panel of previously unidentified protein antigens, many containing multiple epitopes that were frequently recognized by IgG antibodies in CM patients. A subset of these epitopes were also commonly targeted by IgG in PTMs with CM. We adapted several of these targets into a pooled peptide ELISA that discriminated cases from controls in an independent validation cohort with high accuracy (AUC 0.95). Together, these results demonstrate the power of peptide-based highly-multiplexed serology to discover novel fungal antigens, and successfully identify CM specific epitope-resolved targets for new diagnostic assays.