Short-term memory confusions tend to resemble errors made in auditory signal identification tasks. Two kinds of explanation have been advanced to account for this. One postulates an acoustic, the other an articulatory mechanism. Both are compatible with available data. This paper reports a test of a version of the acoustic hypothesis. Auditory imaging was postulated as the mechanism involved. A group of Ss with poor auditory imagery, and a control group, were used in two experiments: a visual shortterm memory task, and an auditory signal identification task. The confusion matrices from the two experiments were rank correlated. Rhos were significant for both kinds of S, implying that auditory imaging is not involved in short-term memory.