This study explored the perception of major, minor, diminished, and augmented triads in two experiments. In the first, subjects judged whether pairs of these triads were of the same or different chord type. In the second, subject rated the similarity of these pairs of triads. Subjects were grouped by degree of musical training. We looked for effects of several variables on the subjects' judgments: the "musical key" suggested by major and minor triads, whether a triad was presented in root position or in first inversion, and whether the notes of the triad were presented successively (i. e., arpeggiated) or simultaneously. Among the three groups of subjects, only the judgments of those who were highly trained appeared to be influenced by the above factors. For these subjects, judgments were influenced by relative position in the circle of fifths. In addition, judgments appeared to incorporate the perceived degree of consonance or dissonance, rather than relative interval sizes.