In this study, three experiments investigated the ways in which melodic and temporal accents influenced the perceived meter of a short musical selection. Stimuli consisted of excerpts from folk songs scored in 6/8 or 3/4 meter. Subjects were asked to match each excerpt they heard with 1 of 2 metrical drum accompaniments (one in 6/8 and the other in 3/4). In the first experiment, a set of isochronous seven-note folk melodies were presented to subjects, each of which could be interpreted in triple or duple meter. The melodies varied according to predictor variables like contour change, melodic leaps, registral extremes, melodic repetition, and harmonic rhythm. Variables that significantly contributed to metrical perception were local pivots, pattern repetition, note repetition, and tempo. Experiment 2 was a repetition of experiment 1 but with stimuli that were twice as long, attempting to address problems that may have arisen from the short length of the stimuli in experiment 1. Experiment 2 yielded similar results to experiment 1, but contour change seemed to play a more important predictive role, perhaps reflecting an emphasis on global rather than local processing. Experiment 3 introduced non-isochronous melodies but used the same paradigm to see whether temporal and melodic cues interact to determine metrical perception. Results showed that tempo and note duration were the strongest predictors of metrical perception, and melodic accent variables failed to make significant unique contributions in the presence of duration cues. Overall, the findings suggest that listeners integrate multiple features, both melodic and rhythmic, to perceive meter.