To provide a voice evaluation device which can calculate an evaluation value of voice quality in a short period and can obtain an evaluation value of voice quality reflecting a situation of accumulation of transmission delay even if an evaluation value of voice quality is calculated based on transmission delay measured in a prescribed period.

In the voice evaluation device, accumulated delay (Tovr) which is the accumulation of delay of RTP packet transmitted in an evaluation target period (one second) by a transmission side communication terminal is calculated, and transmission delay of RTP packet and accumulated delay are repeatedly used until termination of a prescribed period (60 seconds). The voice evaluation device calculates an evaluation value of voice quality based on transmission delay of RTP packet and accumulated delay.