The study of the autonomous nervous system (ANS) has played an important role, over the last years, in prognostic and diagnostic of cardiac diseases, as well as, in the assessment of psychological stress. The most common techniques to evalute the balance of the ANS are invasive and unable to provide a continuous monitoring of the patients. The advances in technology and the development of wearable sensors have provided new alternative methods to study the ANS. The analysis of Heart Rate Variability (HRV) and Electrodermal Activity (EDA) are nonivasive methods to assess the ANS with wearables devices. The wearable device used provides information about HRV with the acquisition of photoplethysmography signals from the wrist and EDA from the fingers. The processing of the biosignals was performed by submitting the participants to a mental arithmetic stress test. The results showed that the participants exhibited two distinct response during stress - “Flight or Fight”. These responses were classified using machine-learning techniques. The constructed models were able to predict how the subjects will respond in a situation of stress, based only on baseline features. The accuracy of the models using only HRV baseline features was of approximately 80% and the accuracy using simultaneously HRV and EDA baseline features was of 77%, when assigning the correct response during stress to the participant.