A voice retrieval apparatus includes a display, a memory, and a processor executing a voice recording process of storing recorded voices in the memory, an accepting process of accepting a retrieval term, a retrieval process of retrieving, from the recorded voices, a candidate segment where an utterance of the accepted retrieval term is estimated, a replay process of replaying voices in the candidate segment retrieved in the retrieval process, and a display control process of adding a marking to display information indicating a transition of the recorded voices in time based on the replay result of the voices in the candidate segment in the replay process, and displaying the display information with the marking on the display, the marking specifying an utterance location of the voices in the candidate segment.