Both musically trained and untrained adults can reproduce the tempo of familiar music with high precision. However, conflicting evidence exists as to how well representations of tempo are preserved within musical imagery. The present study investigated whether previous conflicting evidence might result from the use of different tasks to measure imagined tempo. Tempo judgments for familiar music were collected in a repeated-measures design using two imagined music tasks and one perceived music task. In one imagined music task participants tapped in time to the beat of the imagined music (Imagery (motor) task), while in the other they did not move in time with the music and instead adjusted a click track to the beat (Imagery (non-motor) task). Overall, performance was most accurate on the perceived music task, in which all musical cues were present. Performance on the Imagery (motor) task was also significantly more accurate than performance on the Imagery (non-motor) task. Training and active engagement with music positively predicted imagery task performance, whereas perceived music task performance was influenced by properties related to the song stimuli, such as familiarity and the original, recorded tempo. Results are discussed in relation to previous literature on auditory–motor interactions and musical expertise.