The responses based on equivalence-equivalence relationships have a close resemblance, both in its structure and in the behavioral demands involving with second order conditional discrimination. Traditionally, such discrimination has been used physical relationship, while relations equivalency-equivalence relationships are completely arbitrary. In this paper we compare the performance of 20 subjects in equivalence-equivalence tasks both with the classic, used from the early research in this area (Barnes, Hegarty & Smeets, 1997), as the format of second order conditional discrimination (Moreno, Cepeda, Hickman, Penalosa & Ribes, 1991, for example). Extending the work on second- order conditional discrimination based on arbitrary relationships and assessing the possible transfer between the two formats assessment. Although there is a better implementation of the subjects in the classical equivalenceequivalence tasks, and even some facilitator effect, the results show that the tasks of second-order conditional discrimination may also involve this kind of symbolic relations.