This study presents an account of position saliency in terms of children's ability to utilize graphic information, and in particular the serial encoding of information from letters in a graphic pattern. By varying the number and position of the letters distinguishing graphic patterns (positive condition) in a short-term recognition memory (STRM) task, the relative use of information from different parts of the graphic pattern can be determined. Under STRM, the subject must analyze, remember, and utilize the graphic information from the Target (T) stimulus, in judging whether or not the Comparison (C) stimulus is the same as the T stimulus. In Experiment one, Exposure Time (ET) of the T stimulus, Retention Interval (RI) between T and C stimuli, and the Positive condition were varied. The results indicated that both second semester prereading kindergarten children and first graders are more likely to detect a difference between the T and C stimuli in the initial letter position than at the middle or final position. Experiment two used a similar paradigm with a fixed ET and RI. The results revealed neither a position effect nor any indication that entering kindergarten children's performance changed during the study. (Author/TS)