Wild Tasmanian devil populations have declined precipitously due to the spread of a transmissible cancer, Devil Facial Tumor Disease (DFTD). DFTD causes tumours on the face, neck, and oral region that results in death from metastases, and inability to compete and feed. Despite a latent period that can be greater than one year, the diagnosis of DFTD is only possible by visual inspection once tumours have appeared. Confirmation of DFTD requires pathological examination of tumour biopsies. In this context, a pre-clinical test to detect infection before tumours appear is critically needed to control the advance of the epidemic and prevent extinction of devils. Major advances have been made in early detection of human cancer from bodily fluids, based on molecular analysis of extracellular vesicles (EVs). EVs are cell-derived, membrane-enclosed structures of 50-300 nanometers that are present in bodily fluids such as blood serum. We used data-independent acquisition mass spectrometry techniques to analyse the proteome of extracellular vesicles derived from blood serum of a discovery cohort of devils (12 DFTD+ and 10 captive healthy control). This analysis revealed that cathelicidin-3 (CATH3) could classify devils with DFTD in late stages from healthy controls with 100% sensitivity and 100% specificity. To validate these results, we expanded this sample to 65 devils in different stages of DFTD (33 DFTD+; 15 latent, sampled 3-6 months before diagnosis; 17 control) and found that CATH3 predicted DFTD status with 87.8% sensitivity and 94.1% specificity. Further, it classified latent devils as DFTD+ with 93% sensitivity and 94% specificity. Cathelicidins are antimicrobial peptides, and while humans possess only one cathelicidin gene, there is evidence that the human CAP18 gene plays a role in cancer regulation. On this basis, we hypothesize that cathelicidin- 3 (CATH3) is involved in the pathogenesis of DFTD and is a very promising candidate biomarker for early DFTD diagnosis. A preclinical biomarker for DFTD will greatly improve the capabilities of management and conservation actions including allowing healthy devil selection for insurance population, enhancing epidemiological monitoring, and improving outcomes of eventual therapeutic or prophylactics deployments.