Abstract This proof-of-principle study investigates the diagnostic performance of the Aeonose in differentiating malignant from benign thyroid diseases based on volatile organic compound analysis in exhaled breath. All patients with a suspicious thyroid nodule planned for surgery, exhaled in the Aeonose. Definitive diagnosis was provided by histopathological determination after surgical resection. Breath samples were analyzed utilizing artificial neural networking. About 133 participants were included, 48 of whom were diagnosed with well-differentiated thyroid cancer. A sensitivity of 0.73 and a negative predictive value (NPV) of 0.82 were found. The sensitivity and NPV improved to 0.94 and 0.95 respectively after adding clinical variables via multivariate logistic regression analysis. This study demonstrates the feasibility of the Aeonose to discriminate between malignant and benign thyroid disease. With a high NPV, low cost, and non-invasive nature, the Aeonose may be a promising diagnostic tool in the detection of thyroid cancer.