Autism spectrum disorders (ASD) are considered as a series of pervasive developmental disorders, and characterized by deficits in social interaction, delays and deviance in communication, and repetitive behaviors, rituals and interests. Emotional empathy refers to unconscious experience of emotions consistent with and in response to those of others and imitation of the facial expression, induced by shared representations of perception and emotional contagion. Eye-tracking studies have demonstrated that the visual fixation patterns of children with autism in social situations differ from normal patterns. Children with autism do not seem to attend spontaneously to salient features of the social stimulus in the way that most people do. It is possible that this reflects an attentional style that limits empathic arousal. Children with autism, relative to typically developing children, focused less on the eye region of faces when viewing social stimulus, because they avoid eye contact with others in order to decrease arousal. Previous studies explored emotional empathic responses to affective stimuli in children with ASD, but their results are inconsistent. The present research was designed to investigate the emotional empathy and the face scanning patterns of children with ASD using biofeedback measurement and eye-tracking. Fifteen children with ASD, fifteen typically developing (TD) children, and fifteen children with intellectual disability (ID) viewed eighty emotional faces with happiness, sadness, fear and anger from the Chinese Facial Affective Picture System (each emotional category consists of twenty pictures). Facial electromyography (EMG) activity, eye movements and automatic physiological responses, including skin conductance (SCR), skin temperature (TEMP) and PLUS, were recorded during the experiment. Comparing with typically developing children, children with ASD demonstrated atypical processing patterns of emotional faces with reduced responses of automatic facial EMG, and automatic physiological (SCRs, TEMP, PLUS). This suggested that a failure to perceive emotional faces in children with ASD. Furthermore, results of eye-tracking showed that the children with ASD less viewed the eye areas than the typically developing children. At the same time, the fixation time and fixation counts of face for children with ASD were significantly less than the intellectual disability group and the typically children group. The fixation time of fear faces for children with ASD was significantly less than the happy faces and sad faces. This suggested that the limited eye gaze for emotional faces in children with ASD. The results suggest that a failure to orient to emotional relevant stimuli (four basic emotional faces: happiness, sadness, fear and anger) may be caused by the lower fixation time to the eye areas, and avoidance of emotional stimulus may become the default coping mechanism for some individuals with autism. In conclusion, the current study demonstrates deficits in emotional empathy and limited eye gaze in children with ASD, which may be contributing to their problems in the face scanning patterns.