In each of three experiments, confusability between members of a parafoveally exposed pair of letters affected accuracy of identifying the peripheral, but not the central, letter. Confusability was determined from a confusion matrix developed for each subject. In Experiment I, only one letter in each pair was identified on each exposure, and the position of pair members was varied over trials while the absolute position of the pair was held at a constant distance from fixation. In Experiment 2, both letters were identified on each exposure. In Experiment 3, the criterion letter was presented at a constant distance from fixation, and both letters were identified on each exposure. Since results in Experiment 8 were the same as in Experiments 1 and 2, the effect cannot be explained with reference to an interaction between confusability and acuity. The implications of the findings for various models of visual information processing are discussed.A number of studies have compared the effect of placing a nontarget on the peripheral side of a parafoveally projected target with placing it on the central (foveal) side (Banks, Bachrach, & Larson, 1977;Banks, Larson, & Prinzmetal, 1979;Chastain, 1981;Chastain & Lawson, 1979;Krumhansl, 1977;Krumhansl & Thomas, 1977;White, 1981). The inclusion of target-nontarget confusability as a variable in these studies has been occasional and largely incidental, with inconclusive results. White (1981) varied the similarity of adjacent borders of parafoveally presented geometric forms. Although performance was poorer when adjacent contours were similar, the magnitude of the effect was not significantly different for central peripheral targets. However, it is difficult to directly relate this finding to the effects of stimulus confusability and target position on performance. Only similarity between adjacent borders, and not confusability between stimuli, was varied. It would also be inappropriate to draw conclusions from the absence of a significant effect. Krumhansl and Thomas (1977) used three sets of letters as stimuli. Inter-and intraset confusability was not examined for those particular letters, but instead was derived from Townsend's (1971) alphabetic confusion matrix. Although some tendency was noted for the difference between performance with confusable and performance with nonconfusable stimuli to be smaller when the nontarget was peripheral than when it was central, the effect was not statistically significant. Krumhansl (1977), using those same sets of letters,The author would like to thank two anonymous reviewers for their helpful comments on an earlier version of this article. Requests for reprints should be addressed to Garvin Chastain, Department of Psychology, Boise State University, Boise, Idaho 83725.found a similar effect. She also found that the difference between the effects of placing the nontarget on the target's foveal side and on its peripheral side was greater with nonconfusable stimuli. Krurnhansl (1977) attributed this result to the mislocalization of featur...