The article presents the problem of the scope and causes of the coherence of three empirically and definitionally distinctive components of an emotional reaction - the experiential, physiological and expressive one. Coherence has been defined as integrated, parallel and emotion-specific changes in respective components of an emotional reaction. The evolutionary theory of emotions postulates different response patterns for every emotion and assumes coherence between emotional components, underlied by neurological mechanisms. However, the review of literature on the subject indicates that the most frequent research outcome has been the lack of or only partial coherence between emotional components. The origin of this lack of coherence may be sought in individual differences with regard to emotionality, i.e. in certain personality dimensions or traits. The conducted research was supposed to determine if styles of coping with threatening stimuli, described by D. Weinberger, G. Schwartz, R. Davidson (1979), condition the level of coherence between the described components of an emotional reaction: the experiential, expressive and physiological one. 60 women took part in the study - 20 repressive, 20 high-anxious and 20 low-anxious. The participants' task was to give a public speech about the most liked and the most disliked trait of their personality. The research instruments and procedures were selected according to the possibility of gaining the record of the emotional process changing in time. Along with the physiological method - ECG Holter - the author applied the Facial Action Coding System - FACS - of P. Eckman, W. V. Friesen and J. C. Hager (2002). The results indicated that there were significant differences in the experiential, expressive and physiological responses to a social stress between repressive, high-anxious and low-anxious participants.