Studies comparing the effectiveness of the stimulus-pairing-observation and matching-to-sample procedures in facilitating equivalence relations have reported conflicting findings. This study compared the effectiveness of these procedures and examined the effect of stimulus arrangement and the number of training trials completed prior to each exposure to tests for symmetry and equivalence. Overall, the matching-to-sample procedure resulted in a greater percentage of participants demonstrating equivalence, and with fewer training trials, than did the stimulus-pairing-observation procedure. The one-to-many stimulus arrangement was more effective than the many-to-one and linear arrangements, overall. However, there was an interaction between the type of training procedure and stimulus arrangement. Participants who completed 120 training trials prior to each test were more likely to demonstrate equivalence than participants who completed 60 trials. This appeared to be the result of completing a greater number of trials prior to each test rather than of the number of training trials completed overall. (Contains 2 tables and 1 figure.)