Three students with moderate disabilities were taught to read and match-to-sample sight words comprising stimulus sets based upon the four food groups. We taught students conditional discriminations within four four-member sets using a single-sample/four-comparison procedure. Students were taught A-B, B-C, and C-D conditional discriminations for each of the four potential stimulus classes. Subsequent probes tested for relations based upon symmetry and one-node and two-node transitivity. The performances for all students indicated that symmetric relations emerged before one-node transitive relations, and that one-node transitive relations emerged before two-node transitive relations. These results are consistent with a pattern of responding, referred to as a "nodality effect," in which relations with fewer nodes are demonstrated prior to the demonstration of relations with a greater number of nodes. These results extend this area of research to sight-word reading for students with moderate disabilities.DESCRIPTORS: stimulus control, stimulus equivalence, nodality effects, restricted stimulus control, students with moderate disabilities Equivalence relations are defined by the demonstration of specific properties among stimuli that do not share physical similarities. The relations de- (Saunders & Green, 1992;Sidman & Tailby, 1982). When each of these properties is demonstrated among a set of stimuli, an equivalence dass is established. In laboratory settings, following the training of a set of unidirectional match-to-sample relations, initial testing often indicates the presence of the defining relations ofequivalence (Sidman, 1992