A series of experiments examined the precursor to the relational evaluation procedure (pREP). The pREP is capable of facilitating derived relational responding, but it is less effective than matching-to-sample (MTS) in producing equivalence class formation. Part 1 of the present study tested possible reasons for the inferiority of the. pREP relative to MTS. The first two experiments compared the performances of subjects on two modified versions of the pREP with their performances on a simultaneous MTS procedure. The modifications did not improve performances on the pREP. Experiment 3 compared the pREP with a delayed MTS procedure but again MTS was more effective than the pREP in producing equivalence. Part 2 of the study determined whether pREP equivalence responding could be facilitated by preexposing subjects to a history of MTS training and testing. In Experiment 4, subjects were trained and tested on a MTS procedure until they reliably produced both symmetry and equivalence, and were then exposed to pREP training and testing using the same stimuli, and relations among stimuli, as employed for the MTS procedure. Following this, subjects were exposed to pREP training and testing using novel stimuli. All subjects reliably produced both symmetry and equivalence responding on the pREP with both familiar arid novel stimuli. Experiment 5 determined whether using the same stimulus sets across the two procedures was necessary. Results showed that if symmetry and equivalence were shown on a MTS procedure followed immediately by pREP training and testing using novel stimuli, only 2 out of 4 subjects successfully demonstrated both symmetry and equivalence using the latter procedure.