In two experiments, we examined the effects of manipulating the density of stimuli on comparison difficulty in a comparative judgment task. In Experiment 1, subjects were slower at judging the relative size of a pair when the members were adjacent items in the linear order than when the members were separated by items of intervening magnitudes. In Experiment 2, the advantage of choosing the larger rather than the smaller of two large stimuli (e.g., the congruity effect) increased when the linear order included many small items. In contrast, the advantage of choosing the smaller of two small items increased when the linear order included many large items. The applicability of the range-frequency theory (Parducci, 1965) to these results is discussed.One of the more interesting effects of context is that it can qualify the truth of statements about even the most basic properties of objects. For example, the statement "the moon is small" is almost certainly false in the context of statements about everyday objects, but is probably true in the context of statements about other celestial bodies in the solar system.For some time, researchers have known that context has similar effects on relational judgments (Banks, 1977; Cech & Shoben, 1985). Relational or comparative judgments involve comparing two objects on some specified dimension. For example, elephants are larger than fleas (Cech & Shohen), lions are fiercer than sheep (Holyoak & Mah, 1981), and 2 is less than 5 (Banks, Fujii, & Kayra-Stuart, 1976). These results can he quite striking. For example, Cech and Shohen found that the size and direction of the congruity effect for selected pairs changed as a function of context. The congruity effect is the finding that it is relatively easier to discriminate two objects when the form of the question matches the magnitude of the objects. For example, it is easier to choose the larger rather than the smaller of elephant and rhino, but it is easier to choose the smaller rather than the larger of mouse and rat. Initially, their subjects compared the sizes of animals ranging from that of a flea to an elephant. After one block of trials in this full range, these comparisons were subsequently restricted to a smaller range of items: Preparation of this article was supported in part by PSC/CUNY Faculty Research Award 661486 to the first author. We wish to thank Claude Cech, Edward Shoben, Allen Parducci, and two anonymous reviewers for their comments and suggestions on a draft of this article. Requests for reprints should be sent to K. Sailor, Department of Psychology, Lehman College, Bronx, New York 10468-1589. from flea to beaver or sheep to elephant. In the full range, Cech and Shohen found that small pairs (such as rabbit-beaver) were discriminated more rapidly under the instruction to select the smaller item, and large pairs were discriminated more rapidly under the instruction to select the larger item. However, in the restricted range, the larger of the small pairs (such as rat-rabbit and rabbit-beaver) were discrimina...