<p>This study investigates whether physiological signals, specifically electrodermal activity (EDA) and photoplethysmography (PPG), can be used to detect belief and truth judgments during misinformation exposure. In a controlled experiment, participants evaluated climate-related claims while their EDA and PPG signals were recorded. Each trial was labeled with the objective veracity of the claim and the participant’s belief, enabling three classification tasks: binary veracity detection, belief classification, and a joint belief-veracity classification. Using handcrafted features and multiple machine learning models, results show that EDA yields higher classification accuracy than PPG. However, performance declines in the joint classification task, underscoring the complexity of modeling interactions between belief and truth. This work introduces a novel dataset and highlights the potential of biosignals in misinformation detection.</p>