The purpose of this experiment was to replicate and extend previous work which showed substantial differences between good readers (GR) and poor readers (PR) in the time taken to encode single words. The technique used was based on the memory-scanning and visual-scanning procedure first used by Sternberg. The subjects for the study consisted of 30 fourth graders and 30 sixth graders. Of these, 15 subjects were selected for being the poorest readers in their class, and the other 15 were selected as being the best readers in their class on the basis of their scores on the Wide Range Achievement Test. Stimuli were 10 picutres of common objects chosen from the Stanford-Binet picture vocabulary test and their printed word equivalents. For half of the subjects, all of memory set (MS) size two (two items in memory) was run before MS size four; for the other half, this was reversed. The subjects viewed slides of the pictures or words and indicated whether the slides were the same or different. The results showed no differences between GRs and PRs or between grades in preference or in pictorial versus word mode of encoding. However, the substantial reading ability and grade differences found in this and previous studies indicate reader ability difference in the facility to encode single words. (WR)