To test the hypothesis that memory for pictures is less dependent on rehearsal than that for words, two experiments were performed by the use of a variation of Brown-Peterson paradigm. Experiment I examined the degree in the decrement produced by the rehearsal-preventing tasks. The results revealed that memory for pictures was as vulnerable to interference from the rehearsal-pre-venting tasks as that for words. This was inconsistent with the hypothesis. Experiment II replicated Experiment I with additional conditions. The results also showed that memory for pictures is dependent on rehearsal to the same extent as that for words. These results were inconsistent with the hypothesis.