The detection superiority of perceived direct gaze has been considered as a general phenomenon, but several recent studies had found that the advantage effect for direct gaze detection would disappear under the frontal-viewed condition. Recent studies showed that the perception of gaze includes two periods: the component information processing period and the configural information processing period. Accordingly, there are two possible reasons to explain why direct gaze is not significantly faster than averted gaze under the frontal-viewed condition: (1) The detection-time of direct gaze and averted gaze is equally fast in configural information processing, having a cancelling effect on each other, so the detect direct gaze is not faster than the averted gaze. (2) Detecting direct gaze is faster than averted gaze in the configural information processing, but detecting averted gaze is faster than direct gaze in the component information processing. Thus, both processings cancelled each other out in this condition. So, what was the real reason behind the disappearance of the detection superiority of perceived direct gaze? Previous studies were not clear about this issue, we intended to investigate this question in the current study. We had conducted three experiments in this study as follows: Experiment 1: We used a visual search task. Stimuli pictures are consisted of the eye region cut from color photographs of one Chinese female face. There were two gaze directions been used in our experiment: direct gaze (the eye directed straight towards the camera/observer), and averted gaze (the eye averted by 30°).This experiment was a two by two factor design with the target gaze direction (direct gaze and averted gaze) and the display orientation (upright and inverted) as independent factors. Fourteen volunteers (7 males, 7 females) participated in the experiment. They were asked to judge whether there was a discrepant eye in the array as quickly and accurately as possible. The result demonstrated that detecting direct gaze was not significantly faster than detecting averted gaze under the upright condition, while detecting averted gaze was notable faster than detecting direct gaze under the inverted condition. Experiment 2: The experiment task and stimuli were set as the same as in experiment 1 except that two different gaze deviated angle (10°, 20°) were adopted as averted gaze stimuli. The experiment included three factors: target gaze direction (direct gaze, averted gaze), gaze deviation (10°, 20°) and display orientation (upright, inverted). Fourteen volunteers (7 males, 7 females) participated in this experiment. The result revealed that the detection advantage of direct gaze only occurred when gaze deviated 10°in the upright display condition; the reaction time (RT) for detecting averted gaze was significantly faster than that for detecting direct gaze when gaze deviated 20° in the inverted display condition, while there was no significant difference between the RT for detecting direct gaze and the RT for averted gaze in gaze deviation 10° condition. Experiment 3: The experiment task was the same with the experiment 1, however, the face picture was used as stimuli. As for the averted gaze stimuli, three different gaze deviations (10°, 20°, 30°) were adopted. There are three factors in this experiment: target gaze direction (direct gaze, averted gaze), gaze deviation (10°, 20°, 30°) and display orientation (upright, inverted). Sixteen volunteers (8 males, 8 females) participated in the experiment. The result showed that when face picture was displayed upright, the detection advantage of direct gaze was observed under three gaze deviated angle conditions; when face picture was displayed inverted, the RT for detecting direct gaze and averted gaze was not significantly different under all gaze deviated angle conditions. The above results showed that the disappearance of advantage for direct gaze detection was probably caused by the specificity of averted gaze in the component information processing. The detection superiority of perceived direct gaze is a universal phenomenon, however, it is vulnerable to be affected by the component information in such kind of laboratory measurement.