Talk therapy such as talking about one’s feelings and problems has been thought to be an effective method for minimizing the impact of negative emotional events on current experience. Recently, a number of studies of affect-labeling / matching were conducted on healthy individuals, and the neural mechanisms by which putting feelings into words may alleviate negative emotional responses have been discovered. After going through a major traumatic event, people might suffer from mental disorder to varying degrees, some people even develop post-traumatic stress disorder (PTSD). And talk therapy is often used as one of the psychological intervention methods for PTSD. However, some researchers believed that using talk therapy such as critical incidence stress debriefing can worsen rather than ameliorate later trauma symptoms. Different clinical psychologists have different opinions, and the validity of talk therapy for PTSD has become one particular bone of contention. Therefore, the present study aims to explore this problem by using affect labeling paradigms. The subjects in the present study included three groups: the PTSD group who experienced the earthquake, the traumatized control (TC) group who experienced the earthquake but without PTSD and the non-traumatized control (NTC) group who never experienced any traumatic events. All the subjects performed an affect-labeling / matching task firstly and then made an odd-even decision to a digit that used a picture as background. The background pictures might be either the “target picture” which was affect-labeled or affect-matched in the preceding task or a new picture which was similar with the “target pictures”. Experiment 1 used earthquake pictures as materials and showed that the PTSD individuals responded significantly fast and accurately when the affect-labeling pictures were in the background, however, the NTC individuals responded significant accurately when the new pictures were in the background. Experiment 2 used negative emotional faces as materials and showed that the PTSD individuals responded significantly fast when the affect-matching pictures were background. However, the NTC individuals responded significantly fast when the affect-labeling pictures were background and responded significantly slow when the affect-matching pictures were in the background. These results suggested that after the affect-labeling task, PTSD individuals would inhibit the emotion of the trauma-related earthquake pictures, and it also provided scientific evidence that talk therapy is an effective treatment and intervention method for PTSD individuals. In addition, the inhibition effect after affect-matching of the negative emotional faces indicated that the PTSD individuals had the avoidance tendency to trauma-unrelated negative emotional faces.