Musical intervals, the elementary tonal relations of music, are thought to have widely disparate psychological effects. However, in previous empirical investigations of this proposition, effects due to musical intervals have been routinely confounded with those due to the frequencies, frequency differences, and mean frequencies of the component tones of the interval. An experimental design that can provide an unconfounded musical-interval effect was employed in the present study, in which ratings for 14 musical intervals formed at two different mean frequencies were collected from 16 subjects. Although the results indicated that certain musical intervals do indeed differ in their psychological effects, 7 of the 14 intervals were never discriminated from one another on any of the 11 rating scales, and the number of pairwise discriminations observed among intervals was considerably less than the maximum number possible. Thus, although the present results provide the first rigorous demonstration that the musical interval factor can affect verbal responding, they provide at the same time only partial support for the popular notion that each musical interval has a unique psychological effect.