This paper presents our approach for emotion recognition based on Electrocardiogram (ECG) signals. We propose to use the ECG’s inter-beat features together with within-beat features in our recognition system. In order to reduce the feature space, post hoc tests in the Analysis of Variance (ANOVA) were employed to select the set of eleven most significant features. We conducted experiments on twelve subjects using the International Affective Picture System (IAPS) database. RF-ECG sensors were attached to the subject’s skin to monitor the ECG signal via wireless connection. Results showed that our eleven feature approach outperforms the conventional three feature approach. For simultaneous classification of six emotional states: anger, fear, disgust, sadness, neutral, and joy, the Correct Classification Ratio (CCR) showed significant improvement from 37.23% to over 61.44%. Our system was able to monitor human emotion wirelessly without affecting the subject’s activities. Therefore it is suitable to be integrated with service robots to provide assistive and healthcare services.