Weexamine some implications of an attentional model designed to explain dimensional contrast. Pigeons were trained to discriminate rectangular forms under conditions that produced positive dimensional contrast, In two experiments, the spacing of training stimuli was manipulat-edin ways that should have changed the allocation of attention, Experiment 1 compared the effect of an increased spacing between negative and positive stimuli with the effect of a decreased overall range of positive stimuli. Both. manipulations increased the quality of discrimination performance, Experiment 2 demonstrated that dimensional contrast continued to occur under conditions that minimized the difficulty of the task. Results from these experiments were consistentwith quantitative predictions of theattentional.model concerning the interaction of discriminability factors and factors related to the allocation of attentional resources.Most theories of stimulus control assume that performance in discrimination tasks is based on the underlying discriminability between stimuli (e.g., Balsam, 1988;D. S. Blough, 1983;Boneau & Cole, 1967;Heinemann & Chase, 1975;Pearce & Redhead, 1993;Spence, 1937), Discriminability is determined by the way stimuli differ in value along specific physical dimensions or by the number and types of features that constitute different stimuli, It follows that discriminability increases as stimuli are increasingly separated along physical dimensions or as the number of shared features among the stimuli declines. One implication of this view of discriminability is what may be called the monotonicity assumption, The monotonicity assumption holds that there is a monotonicrelation between measures of discriminability and measures of performance in discrimination tasks. As discriminability between stimuli increases, actual discrimination performance should also improve.Although the monotonicity assumption is a reasonable starting point, many findings in the conditioning literature are inconsistent with this assumption, One important violation of the monotonicity assumption appears in the phenomenon of dimensional contrast, Dimensional contrast refers to an enhancement of discrimination performance for relatively similar positive (S+)and negative (S-) training stimuli, as compared with more dissimilar stimuli. For example, D. S. Blough (1975) found that pigeons trained on a maintained generalization procedure with multiple wavelength stimuli showed two related effects: positive and negative dimensional contrast. Positive dimensional contrast refers to the finding that the highest responding occurred during S + stimuli near S-on the wavelength Correspondence concerning this article should be addressed to 1.M. Hinson, Department of Psychology, Washington State University, Pullman, WA 99164-4820 (e-mail: hinson@mail.wsu.edu).continuum. Negative dimensional contrast refers to the finding thatthe lowest responding was found for S-stimuli near S + along the wavelength continuum.Dimensional contrast is regularly observed during a...