Procedures typically used in Experimental Analysis of Behavior for studying equivalence classes formation may constitute an experimental model for studying symbolic behavior. Equivalence classes formation and symbolic behavior are alike in the fact that dissimilar elements, once arbitrarily related, become equivalent, that is, they become substitutable concerning to the control of the repertoires related to them. Difficulties in obtaining equivalence classes with non-linguistic organisms have lead some theorists to speculate that equivalence class formation, and possibly symbolic behavior, is a phenomenon dependent on linguistic functioning. Some studies reported in the literature, however, suggest that difficulties to obtain equivalence classes with non-linguistic organisms may be due to procedural failure to effectively establish stimulus control relations planned by the experimenter. This paper presents a brief review of basic concepts on equivalence relations, discusses some of the possibilities for development of non-coherent stimulus control in equivalence studies, and briefly points to studies of alternative procedures to select stimulus control relations that cohere with the experimentally planned relations.