In an investigation of interactions between scales and intervals in music cognition, melodic intervals were judged in three preceding-scale contexts: diatonic, chromatic, and no scale. Musically less trained and highly trained subjects compared standard and comparison intervals using three response categories: smaller, equal, and larger. Standard intervals began with notes B or C and ascended by 100, 150, or 200 cents. Discriminal dispersion was estimated for each combination of standard and comparison intervals, based on the assumption that the bandwidth of subjective equality was constant. The dispersion width and the modal dispersion corresponded to the equality- related and sizerelated aspects of interval judgments, respectively. The size-related aspect was strongly influenced by the size of the standard intervals. The point of balance, which corresponds to the traditional point of subjective equality (PSE), tended to be smaller as the standard interval became larger. It was, however, anchored to the point of musical equality when the standard interval began with the tonic. The equality-related aspect was influenced by the relationship between the preceding scale and the intervals to be judged. The diatonic preceding scale differentiated the intervals by their positions along the scale, that is, a sharp discriminal dispersion was estimated when the judged intervals were congruent with the diatonic scale. Such differentiation was not clearly observed in the chromatic condition. The relationship between these two aspects of interval judgment and the subject's musical ability is discussed.