The peculiarities of emotional context identification by an examinee listening to audio-taped actor's utterances (joyous, sad, frightened, angry, and neutral) were studied in two situations: when the phrases sounded naturally; and when their meaning was impossible to understand, so that the listener had to use only extralinguistic component of the utterance. The examinees' principal ability to identificate emotional context by extra-linguistical component is shown. There are significant individual differences in the probability of right identification.