Recent eye-tracking technology has grown popularity in various field of research and industry. However, it has been a challenge to identify human emotional states using eye tracking method. The purpose of this study is an empirical attempt to apply the eye tracking device and method to investigate the relationship between human emotions and eye-gaze behaviors. In order to achieve this purpose, nine emotional video stimuli such as Amusement, Joy, Neutral, Sad and Fear are used to stimulate two types of emotions which are positive and negative. These stimuli are displayed and emotions are measured using eye- tracker, Tobii TX 300. Fifteen students from Malaysian-Japan International Institute of Technology (MJIIT) are randomly chosen for the study. The results obtained are in the form of fixation duration and pupil dilation. The results are filtered and tabulated to analyze the related behavior. Mean fixation duration and pupil dilation are calculated to identify the video stimuli with the highest count. ANOVA analysis shows that the significance, p-value is 0.43 for fixation duration. Whereas, p-value 0.003 for pupil dilation indicating there is a significant relationship between eye-gaze behavior and emotion. This study might be favorable in usability issues in human-computer interaction (HCI) contexts.