Temporal binding (TB) refers to a subjective contraction of the time that elapses between an action and a delayed sensory consequence of it. The TB effect has been demonstrated primarily in tasks in which a key press triggers a tone after a short delay and in which participants judge the timing of one or both of these events relative to a visual reference (e.g., a rotating clock hand). In the present Experiments 1 and 2, musicians listened instead to an auditory 'clock' (a metronome) and occasionally made a tap that triggered a delayed tone. The task was to judge whether that test tone fell before, on, or after the midpoint of the interval between two metronome tones. In a passive control condition, participants judged test tones but did not tap. The hypothesis was that the test tone would be perceived as occurring earlier in the active than in the passive condition. However, there was no difference in perceptual judgments. Experiment 3 used a visual metronome as the reference but again obtained negative results, despite greater uncertainty of judgments. It is suggested that TB of action consequences to actions does not occur when the reference signal is rhythmic because such a context enables participants (musicians, at least) to perceive and judge the actual time of occurrence of the action-triggered tone.