This study applied a frequency theory to measure the superiority of pictures over words in both discrimination learning and recognition memory tasks. Three groups of sixth grade students were given separate instructions before viewing slides of either common objects or words. The first group (control) was asked to study the items shown, the second group was told to pronounce each word or name of object three times, and the third group was instructed to picture each word or object in their mind. After studying either a picture or word list, the subjects were instructed to respond to each item, guessing if uncertain, by saying the number of times that the item had been presented for study. Results show that, in comparison to words, pictures were judged as having occurred more frequently, resulted in less within-subject judgments, and produced fewer misjudgments on items not presented during study. At the same time, instructions to pronounce or to picture the items during the viewing of slides had no effect on frequency judgments, either as a main effect or in interaction with differing item types. (MAI)