Four subjects were exposed to the four trial types that define stimulus equivalence from the beginning of the experiment. Procedures were designed to identify acquisition dynamics and relate these observations to responding indicative of equivalence class formation. The data show that, for all subjects, the acquisition of training conditional discriminations was correlated with systematic changes in the subjects' selection responses. The results also indicate that the traditional percent correct measures obscure some important information about the subjects' behavior. The data are discussed in the context of the following statements. 1) Subjects' performances on derived trials are not indicative of relations among stimuli at some other level of analysis but are instances of "equivalencing". 2) "Equivalencing" itself can be characterized as changes in the conditional and conditionally discriminative functions of stimuli involved in the experiment. The potential benefits of this preparation are discussed.