Rhythmic movement coordination exhibits characteristic patterns of stability, specifically that movements at 0 degrees mean relative phase are maximally stable, 180 degrees is stable but less so than 0 degrees, and other coordinations are unstable without training. Recent research has demonstrated a role for perception in creating this pattern; perceptual variability judgments covary with movement variability results. This suggests that the movement results could be due in part to differential perceptual resolution of the target movement coordinations. The current study used a paradigm that enabled simultaneous access to both perception (between-trial) and movement (within-trial) stability measures. A visually specified 0 degrees target mean relative phase enabled participants to produce stable movements when the movements were at a non-0 degrees relationship to the target being tracked. Strong relationships were found between within-trial stability (the traditional movement measure) and between-trial stability (the traditional perceptual judgment measure), suggestive of a role for perception in producing coordination stability phenomena. The stabilization was incomplete, however, indicating that visual perception was not the sole determinant of movement stability. Rhythmic movement coordination is intrinsically a perception/action system.