Marshall and George (1983) had subjects rank the relative frequency of occurrence of the letters of the alphabet and found an overall positive relationship between subjects’ rankings and actual frequencies. Previously unreported from that study were the confidence ratings that subjects assigned to those rankings. Subjects were most confident of the letter E, judged to be most frequent, and of the letter Z, judged to be least frequent. They were least confident in their rankings for the letter G, suggesting that rankings of the letter G would be more susceptible to modification by new situational experiences. In the present research, subjects were preexposed to a matrix of letters that varied the frequencies of these three letters in an attempt to determine to what extent these manipulations of situational experiences would affect later alphabet rankings of these letters. As predicted, only the letter G was affected by the variation in the frequencies of its occurrence in the preexposure matrices. The data suggest a useful procedure for determining the development of background frequency from situational experiences.