We examined music education majors' ability to reproduce rhythmic stimuli presented in melody and rhythm only conditions. Participants reproduced rhythms of two-measure music examples by immediately echo-performing through a method of their choosing (e.g., clapping, tapping, vocalizing). Forty examples were presented in melody and rhythm only conditions; half were in duple, half in triple. Performances were analyzed for melody condition (melody, rhythm only), measure (first, second) and subdivision (duple, triple). A significant interaction occurred between melody condition and measure: participants were more accurate in the first measure for the rhythm only condition, yet more accurate in the second measure for rhythms with melody. Another significant interaction occurred between measure and subdivision: decline in accuracy between measures one and two was more pronounced for duple than triple. Results suggest that melody inhibits the recall of initial rhythmic material, yet assists recall of rhythmic material occurring later in an excerpt.