This paper analyzed the acoustic characteristics of the echo question "〇〇san mo?" asked with the emotions of dislike and surprise. When the utterance expressed the emotion of dislike, the duration was longer, and there was a low F0 at the start of speech production. When the utterance expressed the emotion of surprise, F0 was high in the latter half of speech production. It is more difficult to detect acoustic cues for hearing emotion in unaccented speech compared with accented speech. Thus, it is expected that hearing emotion is more difficult for learners of Japanese.