In order to examine the effect of tone location on temporal order discriminations, tones (200‐ms onset‐onset) in six‐tone sequences were either presented all in one spatial location or were alternated between two spatial locations. On each trial a comparison sequence differed from the standard sequence, either such that the tones in each spatial location contained an order change or such that, when the tones were alternated between locations, only one spatial location contained an order change. In a 3IFC task, subjects chose the sequence in which the component tones were in a different order. Discrimination performance remained above chance in all conditions; however, performance was significantly worse when the tones alternated between locations and only one location contained an order change. These results support the hypothesis that temporal relationships between tones presented in the same spatial location are more salient than relationships between tones presented in different locations; they also support the assertion of Warren and Byrnes [Percept. Psychophys.18, 273–280 (1975)] that subjects can discriminate the order of perceptually unrelated tones better than chance, while extending the result of Bregman and Campbell [J. Exp. Psychol.89, 244–249 (1971)] that subjects perform worse when discriminating the order of perceptually unrelated tones than when discriminating the order of perceptually related tones.