Emotion cognition becomes an important issue in the research areas of smart home, health promotion, etc. In general, a set of multiple physiological signals have been used for emotion recognition. In this study, the associated parameters derived from all the multiple physiological signals during emotion detection are analyzed and assessed. Firstly, a stand-alone multiple physiological signals acquisition system is developed. Secondly, the well-known IAPS (International Affective Picture System) is employed to elicit the affective responses of happiness, pleasure, disgust, and fear for subject test. These physiological signals including photoplethysmogram, electromyogram, electrocardiogram, galvanic skin response, and skin temperature are measured simultaneously. After signal normalization, signal preprocessing, feature extraction, and feature selection, nineteen parameters are input to the support vector machine classifier for the parameter significance evaluation during emotion recognition. From the experimental results, some parameters are not good in emotion analysis based on the significance of paired t-test.